GREAT BARRIER REEF EXPEDITION 1928—29 VOLUME VI BRITISH MUSEUM (NATURAL HISTORY) GREAT BARRIER REEF EXPEDITION 1928—29 SCIENTIFIC REPORTS VOLUME VI 1938—1958 TRUSTEES OF THE BRITISH MUSEUM (NATURAL HISTORY) LONDON: 1968 HU’k ,• NAT. H 1 3 DEC :?58 Trustees of the British Museum ( Natural History) 1968 Publication No. 674 Printed in Great Britain by Eyre and Spottiswoode Limited, Her Majesty's Printers, at The Thanel Press, Margate CONTENTS No. 1. THE LARVAE OF THE DECAPOD CRUSTACEA PALAEMONTDAE AND ALPHEIDAE. By R. Gurney. Pp. 1-60 ; 265 text-figs. [Issued 26th February, 1938.] 2. ON PHL YCTAEXACHLAM YS LYSIOSQUILLIXA GEN. AND SP. NOV., A LAMELLIBRANCH COMMENSAL IN THE BURROWS OF LYSIOSQUILLA MACULA! A. By M. L. Popham. Pp. 61-84 ; 21 text-figs. [Issued 25th March, 1939.] 3. MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA. By C. Crossland. Pp. 85-257 ; 56 pis., 1 text-fig. [Issued 22nd October, 1952.] 4. HYDROMEDUSAE. By P. L. Kramp. Pp. 259-322 ; 2 pis., 9 text-figs. [Issued 22nd July, 1953.] 5. THE SERGESTIDAE OF THE GREAT BARRIER REEF EXPEDITION. By I. Gordon. Pp. 323-333 ; 6 text-figs. [Issued 31st January, 1956.] 6. FORAMINIFERA. By A. C. Collins. Pp. 335-437 ; 5 pis., 3 text-figs. [Issued 16th September, 1958.] . A**- / i . BRITISH MUSEUM (NATURAL HISTORY) GREAT BARRIER REEF EXPEDITION 1928-29 SCIENTIFIC REPORTS VOLUME VI, No. 1 THE LARVAE OF THE DECAPOD CRUSTACEA PALAEMONIDAE AND ALPHEIDAE BY ROBERT GURNEY, M.A., D.So. WITH TWO HUNDRED AND SIXTY -FIVE TEXT-FIGURES LONDON : PRINTED BY ORDER OF TI1E TRUSTEES OF THE BRITISH MUSEUM SOLD BY B. Quaritch, Ltd., 11 Grafton Street, Naw Bond Street, London, W.l; Doub & Co., Ltd., 29 Dover Street, London, W.l; Oxford University Press, Warwick Square, London, E.C.A H.M. Stationery Office, London, S.W . 1 and at The British Mused m (Natural History), Cromwell Road, London, S.W.7 1938 [All rights reserved ] Price Five Shillings [ItBUtd 26 IK February , 1938] . THE LARVAE OF THE DECAPOD CRUSTACEA PALAEMONIDAE AND ALPHEIDAE BY ROBERT GURNEY, M.A., D.Sc. WITH TWO HUNDRED AND SIXTY-FIVE TEXT-FIGURES. CONTENTS PAGE I. PALAEMONIDAE. 1a. Palaemoninae. Leander Group. Leander padficus, Stimpson . . . . . . .3 Leander tenuicornis (Say) ........ 4 Palaemonid B.R.I. ......... 6 Palaemon Group. Brachycarpus biunguiculatus (Lucas) ...... 8 Retrocaris ........... 9 Palaemonid D. I ( R . spinosa, Ortm. ?) . . . • • 9 Palaemonid D. II (R. contraria, Ortm.) ...... 12 Discussion of the Larvae of Palaemoninae ...... 14 lB. P ONTONIINAE ........... 15 Periclimenes Group . . . . . . . . . .16 Periclimenes grandis (Stimpson) . . . . . . • .16 P. agag, Kemp ........... 16 P. diversipes, Kemp .......... 17 Harpilius beaupresi (Audouin) . . . . . • • .18 H. gerlachei , Nobili .......... 20 Corallioearis graminea (Dana) ........ 20 Palaemonid R.S. I .......... 21 Palaemonid B.R. II .......... 22 Palaemonid B.R. Ill .......... 23 Palaemonid D. Ill .......... 24 Palaemonid D. IV . . . . . • • • .25 Palaemonid B.R. IV .......... 26 vi. 1. 1 2 GREAT BARRIER REEF EXPEDITION PAGE Discussion of the Periclimenes Group ....... 27 Mesocaris Group .......... 29 Palaemonid B.R. V ......... 30 Palaemonid R.S. II ......... 31 Palaemonid R.S. Ill ......... 32 Palaemonid R.S. IV ......... 33 Palaemonid B.R. VI ......... 34 Cryptoleander Group . . . . . . . . . .35 Palaemonid B.R. VII ......... 36 Palaemonid B.R. VIII ....... 38 Palaemonid B.R. IX , ..... 39 lc. Anchistioidinae. Anchistioides .......... 41 Anchistioides Species I . . . . . . . . .41 Anchistioides Species II ........ 42 II. ALPHEIDAE. Alpheus ventrosus, M. Edw. ........ 44 A. audouini, Coutiere ......... 45 A. pacificus, Dana 1 ........ . 47 Synalpheus triunguiculatus (Paulson) ...... 48 Alpheid R.S. I .......... 51 Alpheid R.S. II . ...... 52 Alpheid B.R. I ......... 53 Athanas djiboutensis , Coutiere ....... 54 Alpheid D. I .......... 55 Anebocaris ancylifer, Coutiere ...... 56 Alpheid ? D. II . . . . . . . . .56 Relation of Palaemonidae to Alpheidae ...... 58 I. PALAEMONIDAE. Our present knowledge of the larvae of the Palaemonidae is very unsatisfactory. We know fully those of five species of Leander ( L . serratus, L. squilla, L. longirostris, L. fabricii, L. paucidens) and of two species of Palaemonetes, so that the characters of these two genera are fully established. Sollaud has also described the development of a number of species of Leander and Palaemon in which the larval stages are more or less suppressed. Of the many genera and species of the Pontoniinae we know only the develop- ment of two species of Periclimenes of the subgenus Ancylocaris (P. calmani, P. americanus). In addition various larvae taken in plankton have been described under the generic names of Retrocaris and Mesocaris. Lastly I have been able to describe the early larvae of Anchistioides antiguensis (1936) which is regarded by Miss Gordon as a member of the Pontoniinae. The following report is based upon a large and very varied series of larvae from the plankton of the Great Barrier Reef, and I am much indebted to Mr. F. S. Russell for the privilege of making use of it. In order to make the account as comprehensive as possible THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 3 I have brought into consideration also material from Cfhardaqa on the Red Sea and also certain specimens from the ” Discovery ” Expedition. From this abimdant material I have selected for description 17 larval forms which can with some certainty be referred to the Palaemonidae, but cannot be definitely identified. Together they illustrate a variety of form and of structure of appendages which is surprising and raises perplexing problems of systematics. With our present groundwork of knowledge of the two genera Leander and Periclimenes it is possible to separate these forms into the two subfamilies Palaemoninae and Pontoniinae, but even so there remains uncertainty in some cases. The close examination of these larvae shows that the summary I have given of the characters of Periclimenes and Mesocaris (1936. p. 623) is inadequate, since there are details in the structure of the appendages which should be taken into consideration. The significance of such differences naturally only becomes apparent when the survey can be extended over a sufficiently wide field. Another point which I have not before fully appreciated is the correct designation of the spines on the anterior part of the carapace in late stages. The homologies of these spines are not always clear. In Leander the anterior angle of the carapace is pointed, and this point may be taken as the pterygostomial spine, which is lost at the last moult, and does not appear at all in Retrocaris ( Palaemon ?). In L. senatus a small spine appears just above and quite close to the pterygostomial point, and is no doubt the branchiostegal spine, which usually persists in the adult and is absent in the adult of Palaemon. It is, however, present in the larva if Retrocaris is rightly referred to Palaemon or Brachy carpus. The antennal spine, which is present in the adult of both genera, does not make its appearance in the larva. In L. paucidens Yokoya regards a small spine which arises in Stage VI as antennal, but it has the position of the spine which I regard as branchiostegal. Apparently the branchiostegal spine is found in all larvae of the Palaemoninae and is lost later in Palaemon and Brachycarpus, while the antennal spine is very rarely developed in larval life. 1a. Palaemoninae. Leander pacificus, Stimpson. (Text-figs. 1-7.) Locality. Common on the tidal reef-flat at Ghardaqa along the mole leading to the laboratory. One female hatched larvae on March 15th. Description. Stage I. Length 2-75-3-3 mm. Rostrum reaching end of peduncle of antennule, serrated at end. Abdominal somite 5 without lateral spines. Telson much wider than long, the three inner pairs of spines about equal in length. Antennal scale long and narrow, segmented at end, the proximal outer seta reduced to a minute hair. Close to the proximal inner seta is a minute tubercle, such as is found more marked in Pontoniids and some Alpheids. Mandible with one large spine between incisor and molar parts. Maxillule, endopod small, rounded at end, with one subterminal seta ; lacinia 2 with four spines. Maxilla, endopod broad, with well-marked basal lobe bearing two setae ; exopod with five setae, 4 GREAT BARRIER REEF EXPEDITION Maxillipede 1 with endopod faintly divided into two segments, basis not protuberant, with few small setae ; exopod with six setae. Maxillipede 2, basis with three small setae ; endopod of three segments ; dactyl with large terminal spine and two smaller spines on either side. Maxillipede 3, dactyl with very large terminal spine ; without strong spines on basis or ischium. Exopod with eight setae. Rudiments of legs 1 and 2 biramous. Colour : General appearance red. Thorax with a row of four dark red chromato- phores with yellow branches ; a double row of red chromatophores on abdominal somites 1-4. Dark olive chromatophores in base of antennae. Text-figs. 1-7. — Leander padficus, Stimpson. Fig. 1. — Stage I, dorsal. Fig. 5. — Stage I, maxilla. Fig. 2. — „ mandible. Fig. 6. — „ maxillipede 1. Fig. 3. — „ antenna. Fig. 7. — „ maxillipede 3. Fig. 4. — „ maxillule. Remarks. This larva is of perfectly normal Leander type, agreeing in all respects with that of such species as L. longirostris and L. squilla. Leander tenuicornis (Say). (Text-figs. 8-16.) Locality. This species is common on the Halophila- bed at Grhardaqa, and in floating Sargassum. Larvae were hatched in the laboratory on 2nd February and 9th March. Description. Stage I. Length 2 2-2 5 mm. Carapace with small median dorsal papilla ; rostrum nearly as long as antennule, with a few small denticles dorsally at end. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 5 Abdominal somite 5 with lateral spines. Telson broader than long, slightly concave, the innermost pair of spines very small. Antemiule, distal segment with a feathered seta and four aesthetes, of which one is very stout and one seta-like. Antennal scale with four distal segments distinct ; two outer and ten inner and apical setae ; endopod -with a long seta and small spine. Mandible with three large teeth on incisor part, and three spines in the hollow between it and the molar part. Maxillule, endopod slender, bilobed at end, and with two setae ; lacinia 2 with four spines. Maxilla, endopod slender, without basal lobe or seta ; exopod with five setae. Maxillipede 1 , endopod of three distinct segments ; coxa very small ; basis remarkably produced, axe-shaped, with six strong spines ; exopod with four apical setae. Text-figs. 8-16. — Leander tenuicornis (Say). Fig. 8. — Stage I lateral. Fig. 9.— „ antenna. Fig. 10.— „ mandible. Fig. 11.— „ maxillule. Fig. 12.- „ maxilla. Fig. 13. — Stage I, maxillipede 1. Fig. 14. — - ,, maxillipede 2. Fig. 15. — ,, maxillipede 3, dactyl. Fig. 16. — ,, telson. Maxillipede 2, basis with two large spines ; endopod of three segments, with strong spines on segments 2 and 3. Maxillipede 3 with strong spine on basis and at position of division between ischium and merus ; dactyl with two long slender apical spines. Exopods of maxillipedes 2 and 3 with six apical setae. Eudiments of legs 1 and 2 present. Colour : General colour of thorax pinkish yellow ; abdomen nearly colourless when chromatophores are contracted. The large chromatophore behind the eye is dark red, with olive-brown branches, and, when the chromatophores are expanded, the general colour becomes red. The chromatophores in the telson and maxillipedes are olive-brown. Stage II. Length 2-8 mm. Carapace with large supraorbital and anterior dorsal spines. Telson with 8 + 8 spines, the middle pair unusually small. 6 GREAT BARRIER REEF EXPEDITION Antennal scale segmented, with 15 setae ; endopod unchanged. Legs 1 and 2 developed, with exopods bearing eight setae ; rudiments of legs 3 and 4 present, very small. Additional chromatophores have appeared in thorax and abdomen. A few specimens moulted to Stage II in the laboratory ; but neither this nor any later stages were found in the plankton. Remarks. While the larva has most of the characters usual in Leander, the mouth-parts differ considerably. The form of the endopod in maxillule and maxilla, and of the basis in maxillipede 1 is quite unlike that of the normal type, and the presence of strong spines on the basis of maxillipede 2 is a feature unknown in Leander, but found again in some Pontoniids. The long slender spines on maxillipede 3 are also seen in some Pontoniids, and not in Leander. Palaemonid B.R. I ( Leander sp. ?). (Text-figs. 17-24.) Locality. Great Barrier Reef. Stations 3, 44, 53. Description. Last stage. Length 9-85 mm. Rostrum straight, rather deep, with large serrated spine at base and a series of very small dorsal, and one ventral teeth. Carapace with large serrated dorsal spine, small Fig. 17. — Last stage, lateral. Fig. 18. — ,, mandible. Fig. 19. — ,, maxillipede 1. Fig. 20.— ,, maxillipede 2. Fig. 21. — Last stage, palp of maxillule. Fig. 22. — ,, end of telson. Fig. 23. — ,, end of antennule. Fig. 24.— ,, maxilla. supraorbital, branchiostegal and pterygostomial spines. Abdomen straight, somite 5 with small lateral spines and rectangular pleura ; somite 6 with lateral spine. Anal spine absent. Telson five times as long as wide, without lateral spines, and with 4+4 terminal spines, the outer pair large. Antennule, segment 1 with small ventral spine, but without outer distal spine ; THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 7 stylocerite small ; otocyst cavity formed and surrounded by small setae, but not covered by extension of the stylocerite ; exopod deeply cleft. Antennal scale with small distal spine ; flagellum longer than scale. Maxillule, endopod small, with two apical setae. Maxilla, exopod large, outer margin fringed with setae ; endopod with basal lobe bearing two setae. Maxillipede 1 , coxa large, with large bilobed epipod ; basis not protuberant, with numerous setae ; endopod of two segments ; exopod widened at base, with numerous setae, 6 apical setae. Maxillipede 2, coxa with small epipod ; basis without spines ; endopod of four segments, without strong spines. Maxillipede 3, dactyl with one strong- terminal spine. Exopods of maxillipedes with numerous setae. Legs 1 and 2 chelate, leg 2 somewhat the larger. Legs 3 and 4 alike, leg 4 with exopod. Leg 5 not much larger than leg 4 ; propod with a series of small spines along inner margin, but without strong terminal spines ; dactyl with basal spine. Text- figs. 25-31. — Leander (last stage). Fig. 25. — L. serratus mandible. Fig. 29. — L. serratus end of antennule. Fig. 26. — ,, palp of maxillule. Fig. 30. — ,, end of telson. Fig. 27. — ,, maxilla. Fig. 31. — ,, leg 5, dactyl. Fig. 28. — L. squilla elegans, maxillipede 1. Pleopods large, with small points on outer side, and setae on inner side of exopods ; appendix interna well developed. Remarks. This larva differs so little from the corresponding stage in European species of Leander that I feel justified in referring it to that genus. I give some figures of the telson and appendages of the last larva of Leander serratus for comparison (Text-figs. 25-31), from which it will be seen that there are differences in telson and some appendages ; but they are not great. It will be noted, however, that the exopod of the antennule in this form is very much farther advanced than it is in the last larval, or even the first post-larval stage of L. serratus. This early bifurcation of the outer flagellum will be seen also in some other larvae described below. Some specimens which can be definitely referred to Leander of the L. serratus type were found in a sample from Barrier Reef Station 38. 8 GREAT BARRIER REEF EXPEDITION Brachycarpus biunguiculatus (Lucas). (Text-figs. 32-38.) Locality. I am indebted to Dr. J. F. G. Wheeler for larvae of this species hatched out at the Bermuda Biological Station. Description. Stage I. Length 2-2 mm. Carapace with small dorsal papilla and long slender rostrum reaching to about two- thirds length of antennular peduncle. There is a minute point representing the pterygo- stomial spine, but hidden by the eye in side view. Abdominal somites without dorsal Text-figs. 32-38. — Brachycarpus biunguiculatus (Lucas). Fig. 32. — Stage I, lateral. Fig. 36. — Stage I, maxilla. Fig. 33. — „ telson. Fig. 37. — - „ maxillipede 1. Fig. 34. — ,, autenna. Fig. 38. — ,, „ 2. Fig. 35. — ,, maxillule. or lateral spines. Telson with nearly straight posterior margin ; innermost spine very short. Antennal scale slender, with five distal segments well marked ; two outer and ten inner and terminal setae. There is a trace of the inner knob characteristic of Pontoniinae and some Alpheidae. Endopod a straight rod, with long seta and small inner apical spine. Maxillule, endopod bilobed at end, with two small setae. Maxilla, endopod with basal lobe bearing two setae ; proximal lacinia large, with five setae ; exopod with five setae. Maxillipede 1 coxa reduced ; basis protuberant, with several small setae ; endopod unsegmented, or faintly three segmented ; exopod with five setae. Maxillipede 2, endopod THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 9 of three segments, with strong spines on segments 2 and 3 ; basis with two small spines. Maxillipede 3. dactyl with one strong apical spine. Exopods of maxillipedes 2 and 3 with six setae. Rudiments of legs 1 and 2 present. Remarks. Although only Stage I was obtained, it is of some interest as this is the first species of the Palaemon group of which the free larva has been seen. As was to be expected, the larva at this stage cannot be distinguished generically from Leander. We have no indication as to what the later stages may be like, but it is quite possible that they may be included in one of the forms of Retrocaris described below. Retrocaris. This larval genus was foimded by Ortmann (1893) for two species, R. contraria and R. spinosa. A third, R. antarctica , was described by Coutiere (1907). Coutiere referred these larvae to the Palaemonidae, but ventured the suggestion that R. antarcticus might be the larva of Campylonotus vagans , the adult of which species was taken at the same station. While Coutiere’s specimen was only 5 mm. long Ortmami’s R. contraria was 16 mm., and Coutiere regarded it as one of those “ giant larvae " which he believed to be due to abnormal development. I have described myself (1924, p. 123) specimens resembling R. contraria of sizes between 9-5 and 19 mm., and have suggested that they may be larvae of Palaemon. The “ Discovery ” collection contains 20 specimens from six stations, evidently belonging to two species. Of one of these I have also 15 specimens from “Atlantis ” Station 1121. Palaemonid D. I ( Palaemon sp. ?) (Text-figs. 39-52.) Retrocaris spinosa, Ortmann ?, 1893, p. 84. Palaemon sp. ?, Gurney, 1924, p. 123. Locality. “Atlantis ” Station 1121. 37° 53' N., 62° 45' W. Description. Length 18 mm. Rostrum longer than scale, with six dorsal and three ventral teeth ; the basal dorsal tooth slightly larger than the rest and serrated ; end of rostrum serrated dorsally. Carapace with two large serrated dorsal teeth and a posterior papilla ; supra- orbital, hepatic and branc-hiostegal spines present, large. In one of the “Atlantis ” specimens, of 17 mm., red lines are still visible on carapace and abdomen, and their position has been indicated in Text-fig. 39. Abdominal pleura large, those of somites 1-4 with small procurved tooth ; somite 3 with very large procurved dorsal process strongly serrated along anterior margin ; somite 5 with large lateral spines and pleura rectangular, or with small spine at angle ; somite 6 one and a half times as long as deep, with large posterior lateral spines. Anal spine absent. Telson four times as long as wide, with two pairs of lateral spines and 5+5 terminal, of which the second pair is the largest. Antemiule with well-developed stylocerite and strong apical outer spine on segment 1 ; segment 1 with ventral spine ; outer flagellum deeply cleft. Antennal scale widest about 10 GREAT BARRIER REEF EXPEDITION middle and more than three times as long as wide ; apical spine very large ; flagellum long, but broken. Mandible without palp, with six rather small spines between incisor and molar parts. Maxillule, endopod unsegmented, with two setae ; lacinia 2 with five or six spines. Maxilla, exopod very large, outer margin fringed with setae ; endopod with basal lobe bearing two setae. Maxillipede 1 with large bilobed epipod ; exopod widened at base, with several setae at this point ; basis large, with straight inner margin ; endopod unsegmented. Maxilli- pede 2, basis with small setae ; epipod large, with small lobe representing podobranch ; endopod of five segments. Maxillipede 3 with two rudimentary gills. One of these gills is foliated and the other a simple papilla, but they are so close together that it is impossible to designate one of them as a pleurobranch ; both, by position, are arthrobranchs. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 11 Legs 1 and 2 chelate, leg 2 slightly the larger. Legs 3 and 4 alike, with exopods, the dactyl with small terminal spine and propod without long terminal spines. Leg 5 reaching forward to end of antennal scale. Coxa 'with large curved spine ; propod with a group of long straight spines, serrated at the end ; dactyl with large spine at base and long terminal spine serrated at end. Lengths of segments : Dactyl and spine, 1-2 mm. ; propod, 2-38 mm. ; carpus, 1-46 mm. ; ischiomerus, 4-0 mm. Pleopocls large, with setae. Text-figs. 39-52. — Relrocaris ( P (daemon ). Fig. 39. — Palaemonid D. I. Fig. 40. — „ Fig. 41. — ,, Fig. 42. — Palaemonid D. I. Fig. 43.— Fig. 44. — ,, Fig. 45. — „ Fig. 46. — „ Fig. 47.— Fig. 48. — ,, Fig. 49. — ,, Fig. 50. — Palaemonid D. I. Fig. 51- Fig. 52.— 11 mm. “ Discovery ’’ Station 709. Heavy lines on carapace and abdomen indicate red markings, telson. legs 1 and 2. 18 mm. “ Atlantis " Station 1121. Antennule. mandible, maxillule. maxilla, maxillipede 1. 2. telson. leg 5. 7‘6 mm. “ Discovery ” Station 708. part of telson. part of leg 5. Remarks. This description is taken from one of the largest specimens. The smallest measures 7-6 mm. and differs in having fewer and smaller teeth on the rostrum (Text-fig. 50) ; the basal tooth, which really belongs to the carapace series, is strikingly larger than those on the rostrum itself. The telson has an additional pair of terminal spines ; legs 1 and 2 are not chelate ; the pleopods are traceable only as small papillae ; the flagella of the antennule are quite small, the outer one not cleft. This is, no doubt, the same species as I have described from “ Terra Nova ” Stations 46 and 49, but in that case the endopod of maxillipede 1 was found to be segmented, and I must assume an error in the shape of the pleura of somite 5 shown in the figure. In these large, opaque specimens such an error is easily made. 14 GREAT BARRIER REEF EXPEDITION Janeiro. It was, however, also taken on the eastern side of the ocean near the Cape Verde Islands, and so wide a distribution suggests that there may be two species included. Discovery ” Station 691 . „ 701 . „ 704 . „ 706 . „ 708 . „ • 709 . „ 711 . Atlantis ” Station 1121 Terra Nova ” Station 46 . „ 49 . Distribution of Retrocaris Larvae. Number of specimens D. I. D. II. 00° 25' S., 29° 56' W. 1 14° 39' N., 25° 51' W. 1 3° 37' N., 29° 14' W. 2 3° 26' N., 32° 08' W. 6 10° 26' S., 34° 54' W. 1 14° 01' S., 36° 30' W. 4 24° 40' S., 41° 30' W. 5 37° 53' N., 62° 45' W. 15 20° 30' S., 36° 30' W.) 18° 51' S., 33° 40' W.) 6 I have attached the name Palaemon to both the species ; but it is probable that one of them may be, or includes, Brachycarpus unguiculatus . Coutiere’s suggestion that Retrocaris may be the larva of Campylonotus can be dismissed, since the gill formula of the oldest specimens is evidently complete, and there is no race of epipods or arthrobranchs on any of the legs. Discussion of the Larvae of Palaemoninae. Having regard to the fact that we know fully the larvae of all the European species of Palaemoninae, and that they agree in almost every detail, one would suppose that the generic characters would be as securely founded as could well be. Nevertheless Yokoya’s description of L. paucidens (De Haan) shows that there may be important differences, since in this species leg 5 becomes very much larger than in the other species, and indeed approaches the disproportionate size seen in Retrocaris. In L. tenuicornis, while the general form in Stages I and II is exactly the same as in the European species, the structure of the maxilla and maxillipedes is very different, and approaches very closely to that found in some Pontoniinae. It is unfortunate that later larvae in which legs 4 and 5 are developed are not known, but the differences in the maxillipedes are so great that one is tempted to suggest that L. tenuicornis should be excluded from Leander. The adult has in fact certain exceptional characters, as has been pointed out by Kemp (1925, p. 302). The colouring is very unusual, and the pigmentation of the eye is such as is “ not often seen in Leander ”. The cornea is marked with bands of dark pigment, and such bands “ are of very frequent occurrence in Periclimenes and other genera of Pontoniinae ”. The sexual difference in length of leg 2 is also a most unusual feature according to Kemp ; and the endopod of pleopod 1 in the male has an appendix interna. So far as is known this appendix is found among Palaemonidae only in Urocaridella gracilis , Palaemonetes THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 15 hornelli and Anchistioides antiguensis .* In my own specimens the palp of the mandible has two segments only, as in L. squilla, L. cubensis and L. semmelinki. When one considers that the genus Palaemonetes is generally accepted as distinct from Leander although it differs only in the absence of the mandible palp and of the pleurobranch on maxillipede 3 it is not unreasonable to claim that the peculiarities of the larva combined with those of the adult would justify the same separation of L. tenuicornis. Summary of the Characters of the Last Larva. Leander. (Excluding L. tenuicornis (Say).) Carapace with one or more dorsal spines ; pterygostomial and branchiostegal spines present, antennal and hepatic absent. Abdominal somite 3 without dorsal spine ; somite 5 with lateral spines, except in L. paucidens ; anal spine present or absent. Telson with two pairs of very small lateral spines, or without them ; apical spines usually 5 + 5, spine 2 large, but spine 1 very small and sometimes absent. Antennule with ventral spine. Antennal scale with spine. Maxillule, endopod bilobed at end. with one seta. Maxilla, endopod with basal lobe ; exopod with setae along outer margin. Maxillipede 1, basis small, not protuberant, without spines. Maxillipede 2, basis not protuberant, without spines. Maxillipede 3, dactyl with one strong terminal spine. Leg 4 with exopocl (except in L. longirostris) ; developing later than leg 5 ; structure as leg 3. Leg 5 larger than leg 4. but usually not much larger ; dactyl with basal spine. Retrocar is ( Palaemon or Brachycarpus ). As above, but differing as follows : Abdominal somite 3 with dorsal spine. Anal spine absent. Telson with 5 + 5 or 6 + 6 apical spines, very slender. Maxillipede 3 basis large, somewhat produced distally. Leg 5 very long, twice as long as leg 4. 1b. Pontoniinae. The first larva of the following species was obtained by hatching in the laboratory at Ghardaqa : f Periclimenes ( Ancylocaris ) grandis (Stimpson). P. (A.) agag, Kemp. P. (A.) diver sipes, Kemp. Harpilius beaupresi (Audouin). H. gerlachei, Nobili. Coralliocaris graminea (Dana). * According to Schmitt (1935, p. 61) it is present in Leander paulensis (Ortmann), which is probably a synonym of L. tenuicornis. It is also found in the male of Gnathophyllum fasciolatum ; in some Atyidae and in Rhynchocinetes (Gordon, 1935). f For the identification of these and other species from the Red Sea I am much indebted to Dr. M. Ramadan. 16 GREAT BARRIER REEF EXPEDITION Conchodytes biunguiculatus (Paulson) was seen, but the eggs were in quite early stages, while Palaemonella tenuipes (Dana) and Periclimenes petitthouarsi (Audouin) had not begun to breed. It is unfortunate that the species of Periclimenes belong to the subgenus of which the larvae are already known, and that late stages of Harpilius and Coralliocaris could not be identified. Pontoniid larvae were rare in the plankton at that time, but four types are described below. The Barrier Reef plankton contains large numbers of Pontoniid larvae in late stages, among which those of Periclimenes predominate. I have made no attempt to examine the whole of this material in detail ; but I have selected for description those forms which show striking characters, or may be supposed to represent types rather than species. The variety in form and structure of appendages is astonishing and baffling. Group I : Periclimenes. Periclimenes grandis (Stimpson). (Text-figs. 54-60.) Locality. Common, together with P. agag, Kemp, on Halophila near the laboratory at Ghardaqa. Description. Stage I. Length 1- 63-1- 8 mm. Rostrum very small and slender. Abdomen straight ; pleura rounded. Telson with posterior margin nearly straight, spine 4 not much longer than 3, and innermost pair very small. Antennal scale with three distinct segments ; two outer and ten inner and terminal setae ; proximal papilla absent ; endopod with long seta and very small terminal spine. Text-figs. 54-60. — Periclimenes grandis (Stimpson). Fig. 54. — Stage I, dorsal. Fig. 58. — Stage I, maxilla. Fig. 55. — antenna. Fig. 59. — ,, palp of maxillule. Fig. 56. — ,, maxillipede 1. Fig. 60. — „ maxillipede 2, Fig. 57. — ,, maxillipede 3. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 17 Endopod of maxillule small, broad, bilobed at end, with one small seta. Endopod of maxilla with very small terminal seta and small basal lobe ; exopod with four setae only. Maxillipede 1 , endopod not distinctly segmented ; basis protuberant ; exopod with four setae. Maxillipede 2, basis with two small spines ; endopod of three segments. Maxillipede 3, endopod very long and slender, apparently of three segments only ; dactyl with two long subequal terminal spines, as in Leander tunuicornis. In P. amerieanus Stage I, as in P. grandis, there are two long apical spines only; In Stage II two small spines appear at the end of the dactyl, and in later stages the long accessory spine is replaced by a very small one, so that the dactyl then bears only one strong apical spine, with three small spines at its base. Exopods of all maxillipedes with four setae. Rudiments of legs 1 and 2 present. Colour : General colour olive ; with some red and yellow chromatophores. No ventral chromatophores in abdomen. Remarks. The first larva of P. agag cannot be distinguished from that of P. grandis, as described above. Colour differences, if any, were not noted. Periclimenes diversifies, Kemp. (Text-figs. 61-66.) Locality. A very small, colourless species found among living corals at Ghardaqa. Description. Stage I. Length 1-6 mm. Rostrum long and slender. Abdomen very much bent at somite 3 ; pleura rounded. Telson as in P. grandis, innermost pair of spines very small. Antenna, exopod with three distal segments ; two outer setae, of which the proximal one is very small ; inner papilla well marked : endopod with long seta and large spine. Endopod of maxillule not bilobed, with one small seta. Maxilla, endopod with long apical seta and without basal lobe ; exopod with five setae. Maxillipede 1, endopod unsegmented ; basis slightly protuberant, with three setae only. Maxillipede 3, dactyl with a single strong terminal spine. Rudiments of legs 1 and 2 present. Text-figs. 61-66. — -Periclimines diversifies, Kemp. Fig. 61. — Stage I, lateral. Fig. 64. — Stage I, maxilla. Fig. 62. — ,, antenna. Fig. 65. — ,, telson. Fig. 63— ,, maxillule. Fig. 66.— „ maxillipedes 2 and 3. VI. 1. 2 18 GREAT BARRIER REEF EXPEDITION Colour : Abdomen and posterior part of thorax colourless ; very large chromatophore behind eye with red centre but appearing yellow in reflected light. Remarks. While P. calmcmi, P. americanus and P. grandis agree in almost every detail, so that it would be difficult to distinguish them, P. diversipes differs so much from all of them that one would suppose it should be referred to a distinct genus or subgenus. Such differences are the very well-developed antennal papilla ; form of endopod of maxillule ; presence of five setae on exopod, and absence of basal lobe on endopod of maxilla ; and terminal spines of maxillipede 3. Harpilius beaupresi (Audouin). (Text-figs. 67-74.) Description. Stage I. Length 2-15 mm. Rostrum slender, extending beyond eye. Carapace broad. Abdomen slightly bent and rather broad, somite 3 rather swollen ; pleura rounded. Telson broad, rather deeply indented, the innermost spines short, and spine 4 slightly the longest. Antenna, exopod with four distinct segments ; two outer setae, of which the proximal is minute ; inner margin with well-developed papilla ; endopod with long apical seta and large spine. Mandible large, with large spine between molar and trifid incisor part. Text-figs. 67-74. — Harpilius beaupresi (Audouin). Fig. 67. — Stage I, lateral. Fig. 68. — ,, telson. Fig. 69. — ,, antenna. Fig. 70. — ,, maxilla. Fig. 71. — Stage I, maxillipede 1. Fig. 72. — ,, mandible. Fig. 73. — ,, maxillule, palp. Fig. 74, — Stage II, lateral. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 19 Maxillule, endopod slightly bilobed at end, with one seta. Maxilla, endopod with apical seta but no basal lobe or seta ; exopod with five setae. Maxillipede 1 , endopod indistinctly three-segmented ; basis with spine and two setae, not protuberant. Maxillipedes 2 and 3 with similar dactyls bearing one strong terminal spine. Exopods with four setae. Rudiments of legs 1 and 2 present. Colour : Thorax yellowish, abdomen very faintly coloured. Chromatophores behind eye with olive and red pigment and fault yellow branches, the red largely hidden by the olive. Stage II. A few specimens moulted in the laboratory without increase in size. Carapace with anterior papilla, but no dorsal spine ; supra-orbital spine small ; anterior angle pointed. Abdomen sharply bent at somite 3 ; pleura rounded. Telson unchanged except for presence of additional small inner pair of spines. Antenna unchanged, except for appearance of a very small seta at base of terminal spine of endopod. Legs 1 and 2 developed, with exopods bearing four setae. Leg 3 a small rudiment. Text-figs. 75-80. — Harpilius gerlachei, Nobili. Fig. 75. — Stage I, dorsal. Fig. 76. — ,, maxillule, palp. Fig. 77. — ,, maxillipede 1. Fig. 78. — Stage I, maxillipede 3. Fig. 79.— „ „ 2. Fig. 80. — ,, maxilla. 18 GREAT BARRIER REEF EXPEDITION Colour : Abdomen and posterior part of thorax colourless ; very large chromatophore behind eye with red centre but appearing yellow in reflected light. Eemarks. While P. calmani, P. americanus and P. grandis agree in almost every detail, so that it would be difficult to distinguish them, P. diversifies differs so much from all of them that one would suppose it should be referred to a distinct genus or subgenus. Such differences are the very well-developed antennal papilla ; form of endopod of maxillule ; presence of five setae on exopod, and absence of basal lobe on endopod of maxilla ; and terminal spines of maxillipede 3. Harfiilius beaupresi (Audouin). (Text-figs. 67-74.) Description. Stage I. Length 2-15 mm. Kostrum slender, extending beyond eye. Carapace broad. Abdomen slightly bent and rather broad, somite 3 rather swollen ; pleura rounded. Telson broad, rather deeply indented, the innermost spines short, and spine 4 slightly the longest. Antenna, exopod with four distinct segments ; two outer setae, of which the proximal is minute ; inner margin with well-developed papilla ; endopod with long apical seta and large spine. Mandible large, with large spine between molar and trifid incisor part. Fig. 67. — Stage I, lateral. Fig. Fig. 68. — ,, telson. Fig. Fig. 69. — ,, antenna. Fig. Fig. 70, — ,, maxilla. Fig. 71. — Stage I, maxillipede 1. 72. — ,, mandible. 73. — ,, maxillule, palp. 74. — Stage II, lateral. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 19 Maxillule, endopod slightly bilobed at end, with one seta. Maxilla, endopod with apical seta but no basal lobe or seta ; exopod with five setae. Maxillipede 1, endopod indistinctly three-segmented ; basis with spine and two setae, not protuberant. Maxilhpedes 2 and 3 with similar dactyls bearing one strong terminal spine. Exopods with four setae. Rudiments of legs 1 and 2 present. Colour : Thorax yellowish, abdomen very faintly coloured. Chromatophores behind eye with olive and red pigment and faint yellow branches, the red largely hidden by the olive. Stage II. A few specimens moulted in the laboratory without increase in size. Carapace with anterior papilla, but no dorsal spine ; supra-orbital spine small ; anterior angle pointed. Abdomen sharply bent at somite 3 ; pleura rounded. Telson unchanged except for presence of additional small inner pair of spines. Antenna unchanged, except for appearance of a very small seta at base of terminal spine of endopod. Legs 1 and 2 developed, with exopods bearing four setae. Leg 3 a small rudiment. Text-figs. 75-80. — Harpilius gerlachei, Nobili. Fig. 75. — Stage I, dorsal. Fig. 78. — Stage I, maxillipede 3. Fig. 76. — ,, maxillule, palp. Fig. 79. — ,, ,, 2. Fig. 77. — „ maxillipede 1. Fig. 80. — ,, maxilla. 20 GREAT BARRIER REEF EXPEDITION Harpilius gerlachei, Nobili. (Text-figs. 75-80.) Description. Stage I. Length 2 mm. General form as in H. beaupresi but differing from it in structure of appendages as follows : Endopod of maxillule with additional small seta. Maxilla, endopod with minute apical seta and well-marked basal lobe ; exopod with four setae only. Maxillipede 1, basis very protuberant, without spines. Maxillipede 3, dactyl bearing a pair of long subequal spines. Colour : Almost colourless. Dark red behind eye, and orange in stomach region. No chromatophores seen in abdomen, but a small pair in telson. Remarks. While this species agrees with typical Periclimenes in form of maxilla and maxillipedes 1 and 3 so closely that it could not be separated generically, H. beaupresi differs in exactly the same way from H. gerlachei as P. diversipes differs from the other species of Peri- climenes. Coralliocaris graminea (Dana). (Text-figs. 81-89.) Locality. Common among the branches of Stylophora at Ghardaqa. The animal snaps its large claws in exactly the same way as an Alpheid does. The eggs are borne on legs 1-3, in which the coxa is greatly elongated. Leg 1 seems to carry the most eggs, but in each leg there are only two groups of ovigerous setae, two setae at the distal and three at the proximal end of the basis. Stage I. Length 2-2 mm. Body slightly bent at junction of thorax and abdomen, and at abdominal segment 3. Rostrum slender, extending beyond eye. Abdominal pleura rounded. Telson broad, slightly indented, the innermost pair of spines short ; spine 4 as long as spine 2. Text-figs. 81-89. — Coralliocaris graminea (Dana). Fig. 81. — Stage I, lateral. Fig. 86. — Stage I, antenna. Fig. 82. — ,, maxillule. Fig. 87. — ,, antenna, Stage II, tip of endopod. Fig. 83. — „ mandible Fig. 88. — ,, maxillipede 2. Fig. 84. — „ maxilla. Fig. 89. — Stage II, lateral. Fig. 85. — „ telson THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 21 Antennal scale with three segments distinct ; proximal outer seta reduced to a minute hair ; endopod with long seta and large spine. Inner marginal papilla well developed. Maxillule, endopod with two very small setae. Maxilla, endopod slender, with long apical seta but no basal lobe or seta ; exopod with five setae. Maxillipede 1, basis not protuberant, with one spine and a seta ; endopod unseg- mented, or faintly three-segmented ; exopod with four setae. Maxillipede 2, basis with one small spine. Maxillipede 3, dactyl with one strong terminal spine ; exopod with four setae. Large rudiments of legs 1 and 2. General colour brick-red. Stage II. A few specimens moulted to Stage II, but without increase in length. Carapace without supraorbital spine. Abdomen slightly bent ; pleura rounded. Antenna unchanged, but endopod with small seta at base of terminal spine as in Harpilius beaupresi. Legs 1 and 2 developed ; dactyl with strong apical spine. Small rudiments of legs 3 and 4. Remarks. Except in colour, and absence of supra-orbital spine in Stage II, this larva is almost identical with that of Harpilius beaupresi ; indeed the latter resembles C. graminea very much more than that of H. gerlachei. Palaemonid R.S. I ( Periclimenes ). (Text-figs. 90-96.) Locality. Ghardaqa, plankton. Description. Stage V ? Length 5-0 mm. Text-figs. 90-96. — Palaemonid R.S. I. ( Periclimenes .) Fig. 90. — Stage Y ?, lateral. Fig. 94. — Stage V ?, dactyl of leg 4. Fig. 91. — ,, maxilla. Fig. 95. — ,, telson. Fig. 92. — ,, maxillipede 3. Fig. 96. — ,, mandible. Fig. 93. — „ maxillule. 22 GREAT BARRIER REEF EXPEDITION Rostrum shorter than eye, without teeth ; carapace without supra-orbital spines but with pterygostomial angle pointed. Abdomen nearly straight, somites 5 and 6 with lateral spines. Telson with one pair of lateral spines and 4 + 4 distal, the outer pair large. Anal spine absent. Antennule with rudiment of stylocerite, without ventral spine on segment 1 ; outer flagellum unsegmented and not cleft. Antenna, scale narrow, 4| times as long as wide ; flagellum more than half length of body. Maxillule, endopod not bilobed, with two apical setae. Maxilla, endopod with minute apical seta and small basal lobe ; exopod fringed with hairs, but without setae on outer margin. Maxillipede 1, basis protuberant, without spines ; endopod not distinctly segmented. Maxillipedes 2 and 3, dactyl with one large apical spine. Exopods with four setae only. Legs 1 and 2 chelate, nearly equal. Leg 4 longer and stouter than leg 5, without exopod ; propod with long apical spines. Leg 5 long and slender, dactyl continuous with apical spine and without basal spine. Pleopods small, without setae. Colour : Almost colourless ; end of antennal flagellum yellow. Palaemonid B.R. II ( Periclimenes ). (Text-figs. 97-101.) Locality. Barrier Reef Station 38. Description. Stage V ? Length 4-0 mm. Rostrum short, broadening at base, with median ridge and one small dorsal spine ; carapace with supra-orbital and pterygostomial spines, but no antennal. Abdomen nearly straight, somites 5 and 6 with lateral spines. Telson four times as long as wide, with two pairs of lateral spines and 6 + 6 terminal, of which spine 2 is the largest. Anal spine absent. Text- figs. 97-101. — Palaemonid B.R. II. ( Periclimenes .) Pig. 97. — Stage V ?, lateral. Fig. 100. — Stage V maxillule. Pig. 98. — ,, maxilla. Pig. 101. — ,, part of telson. Fig. 99. — „ maxillipede 1. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 23 Antennule, stylocerite small, no ventral spine on segment 1. Antennal scale four times as long as wide, with small apical spine. Maxillule, endopod slightly bilobed, with three setae. Maxilla, palp with terminal seta and well-defined basal lobe ; exopod very narrow proximally, where it bears one large seta only ; outer margin without setae. Maxillipede 1, basis protuberant, without spines ; endopod two-segmented ; exopod with four setae. Maxillipede 3, dactyl 'with one strong terminal spine ; exopod with six setae, the proximal pair very small. Legs 1 and 2 chelate, nearly equal, dactyl equal to, or shorter than, palm. Leg 4 without exopod, very elongated, and slightly longer than leg 5 ; propod with strong terminal spines. Leg 5 long and slender, without propodal spines, the terminal claw without a basal spine. Pleopods small, without setae. Palaemonid B.R. Ill ( Periclimenes ?). (Text-figs. 102-106.) Locality. Barrier Reef Station 62. Description. Stage V ? Length 5-35 mm. Rostrum long and straight, without teeth ; carapace with small supraorbital, bran- chiostegal and pterygostomial spines. Abdominal somite 3 rather protuberant ; somites 106 Text-figs. 102-106. — Palaemonid B.R. III. Fig. 102.— Stage VI ? Fig. 105.— Stage VI ?, leg 4. Fig. 103. — ,, maxillipede 1. Fig. 106. — ,, telson. Fig. 104.— „ „ 2. 5 and 6 with lateral spines. Anal spine absent. Telson nearly four times as long as wide, with two pairs of lateral and 5 + 5 apical spines, spine 2 the largest. Antennule with small stylocerite ; exopod with very small slender extension beyond the thickened sensory basal part. Mandible with three large movable spines. Maxillule, endopod narrow, with two distal setae. Maxilla, endopod with small distal seta and basal lobe ; exopod large with several setae at proximal end, but none on outer margin. Maxillipede 1 , basis very protuberant, without spines ; endopod unsegmented ; exopod with outer basal seta and with four apical. Maxillipede 2, basis protuberant, with two large spines ; endopod of three segments ; epipod absent. Maxillipede 3, dactyl with one strong apical spine. 24 GREAT BARRIER REEF EXPEDITION Leg 4 as long as leg 5, without exopod ; propod with long terminal spines. Leg 5 long and slender ; apical claw without basal spine. Pleopods rather large, without setae. Palaemonid D. Ill ( Periclimenes ?). (Text-figs. 107-112.) Locality. “ Discovery ” Station 701. 14° 39' N., 25° 51' W. Description. Length 8-2 mm. Rostrum slender, as long as eye, with one dorsal tooth at base. Carapace with small supra-orbital, antennal and pterygostomial spines. Abdominal somite 3 rather protuberant ; pleura of all somites rounded. Anal spine absent. Telson four times as Fig. 107. — Stage VI ? Fig. 110. — Stage VI ?, maxillule, palp. Fig. 108. — ,, maxillipecle 2. Fig. 111. — ,, antenmile, end. Fig. 109. — ,, maxilla. Fig. 112. — ,, part of telson. long as wide, with one pair of small lateral spines near middle, and 4 + 4 apical, the outer pair very large. Antenmile with small stylocerite but no ventral spine ; exopod slightly cleft. Antennal scale with small distal spine ; flagellum very long. Maxillule, endopod with apical and two subapical setae. Maxilla, endopod small, with seta in place of basal lobe ; exopod very large, with small setae on outer margin. Maxillipede 1 , endopod short and stout, of two distinct segments ; basis very protu- berant, without spines. Maxillipede 2, basis protuberant, with two very large spines ; endopod stout, of three segments. Maxillipede 3, dactyl with single strong terminal spine. Legs 1 and 2 chelate, leg 2 much the larger. Leg 4 slightly longer than leg 5, without exopod ; propod with long terminal spines. Leg 5 slender ; dactyl without basal spine. Pleopods fairly large, without setae. This species very closely resembles species B.R. Ill, in form of abdominal somite 3, maxillipede 2, etc., and no doubt belongs to the same genus. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 25 Palaemonid D. IV (Periclimenes ?). (Text-figs. 113-118.) Locality. “ Discovery ” Station 1374. 31° 46' S., 29° 46' E. Description. Length 10-34 mm. Rostrum longer than antennule, with one large dorsal tooth ; carapace with dorsal, supra-orbital and pterygostomial spines. Abdominal somite 3 with large straight dorsal spine : somites 5 and 6 with lateral spines ; anal spine absent. Telson four times as long as wide, with one pair of lateral spines and 4 + 4 distal. The outermost spines are very large and fused with the telson, forming a fork within which are six small spines. Text- figs. 113-118. — Palaemonid D. IV. Fig. 113. — Stage VI ? Fig. 116. — Stage VI ?, maxillipede 1. Fig. 114. — ,, telson. Fig. 117. — ,, maxillule. Fig. 115. — ,, maxilla. Fig. 118. — ,, mandible. Antennule with stylocerite, but no ventral spine ; exopod not cleft, but with two small distal segments distinct. Antenna, basis with stout spine ; scale with very large terminal spine ; flagellum very long. Mandible with four movable spines. Maxillule, endopod slender, curved, with two apical setae. Maxilla, endopod with trace of basal lobe and two setae on inner margin ; exopod large, without setae on most of the outer margin. Maxillipede 1 , basis protuberant without spines ; endopod unsegmented ; exopod with outer basal seta and four apical. Maxillipede 2, basis with two very strong spines. Maxillipede 3, dactyl with single large apical spine. Legs 1 and 2 chelate, leg 2 very much larger than leg 1 . Leg 4 as long as leg 5, slender, without exopod ; with long spines at end of propod. Leg 5 long and slender, the terminal claw without basal spine. Pleopods small. 26 GREAT BARRIER REEF EXPEDITION Palaemonid B.R. IV. (Text-figs. 119-122.) Locality. Barrier Reef Stations 16, 43, 50 and 53. Description. Length 6 mm. Rostrum rather deep, reaching beyond eyes, with five small dorsal teeth. In a specimen about to moult to post-larval there are four above and one below under the skin. Carapace with supra-orbital, branchiostegal and pterygostomial spines. Abdominal somites 5 and 6 with lateral spines. Anal spine absent. Telson more than four times as long as wide, with two pairs of lateral spines and 4 + 4 apical, the outer pair very large. Antennule with small stylocerite and no ventral spine. Antennal scale with large spine ; flagellum much longer than scale ; basis with small spine. Fig. 119. — Last stage, lateral. Fig. 121. — Last stage, maxilla. Fig. 120. — ,, maxillipede 1. Fig. 122. — ,, telson. Maxillule, endopod slender, with two long apical setae. Maxilla, endopod small, with small apical and basal setae, basal lobe not distinct ; exopod large without setae on outer margin. Maxillipede 1 , basis protuberant, without spines ; endopod unsegmented ; exopod with one outer basal seta ; epipod not distinctly bilobed. Maxillipede 2, basis protuberant, with two spines. Maxillipede 3, dactyl with one strong apical spine. Legs 1 and 2 chelate, leg 2 slightly the larger. Leg 4 without exopod, very much longer than 3 and 5 ; propod with long apical spines. Leg 5 about as long as leg 3, propod and dactyl without spines. Dactyls of legs 3-5 of post-larval stage, seen under skin, apparently simple. Pleopods large, without setae. Remarks. This form is included in the Periclimenes series by reason of the great elongation of leg 4 and the general structure of the mouth-parts ; but the reduction of leg 5 as compared with leg 4 is a feature hitherto unknown among the Palaemonidae. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 27 Discussion of the Periclimenes Group. It has been pointed out above (pp. 18, 20) that the larvae of Periclimenes diversifies and Harpilius gerlachei indicate quite a different generic grouping of the adults from that which is adopted by Kemp. Comparison of the antennule and mouth-parts of such species as are available to me seem to confirm the conclusion drawn from the larvae. The three species of Periclimenes , P. grandis. P. agag and P. americanus, agree in almost every detail, the only exception being that I find a podobranch on maxillipede 2 only in P. grandis. On the other hand P. diversifies differs from all of them in almost every appen- dage, and in each case approaches much more nearly to Harpilius beaupresi or to Coral- liocaris. In descriptions of species, or even sometimes of genera, the mouth-parts are often neglected, and as I think some characters of importance in the systematics of the Pontoniinae are available which have not had weight given to them, I give here the facts with regard to these species. Antennule and Mouth-Parts of P. diversipes. (Text-fig. 123a.) Antennule : Stylocerite without the inner expansion covering the opening of the statocyst fomid in other Periclimenes. The opening is also free in H. beaupresi and Coralliocaris. It is covered in U rocaridella gracilis. Mandible : The mandibles are not symmetrical, but in P. americanus, for example, the molar part on the one side has a series of large blunt teeth, and on the other rounded knobs or ridges with small bristles on one of them. In P. diversipes, as in H. beaupresi and Coralliocaris these teeth or ridges are replaced by a more or less horse-shoe-shaped Text-fig. 123. — Antennule and mouth-parts of Pontoniinae. a. Periclimenes diversijpes. b. P. americanus. c. Harpilius gerlachei. D. Ii. beaupresi. i. Antennule. ii. Mandible, iii. Palp of maxillule. iv. Maxilla, v. Maxillipede 1. 28 GREAT BARRIER REEF EXPEDITION ridge bearing a brush-like arrangement of small spines. The difference is very striking. In Urocciridella the mandible is similar to that of Periclimenes. In Periclimenes and Urocardella the incisor part is much larger in proportion to the molar part. Maxillule : In Urocaridella and Periclimenes the palp is bifid at the end, the outer lobe very well defined, whereas in P. diversipes, H. beaupresi and Coralliocaris it is absent. Maxilla : The lacinia is deeply cleft in P. diversipes, as it is in other Periclimenes, and in this respect differs from Harpilius, where it is narrower and not cleft. In Coral- liocaris it is narrow, but bilobed. On the other hand the exopod, which is narrow in front in Urocaridella and Periclimenes, is very broad in P. diversipes, Harpilius and Coralliocaris. Maxillipede 1 : Coxa and basis are separated by a well-marked notch in Periclimenes and Urocaridella, and the basal expansion of the exopod is not very wide, whereas in P. diversipes there is no distinction at all between coxa and basis, and the expansion of the exopod is very wide. In both respects it resembles Harpilius and Coralliocaris. Maxillipedes 2 and 3 : There is no difference in these appendages sufficiently marked to be defined, but P. diversipes is exceptional in having the arthrobranch of maxillipede 3 vestigial, as it is in Harpilius gerlachei. Harpilius beaupresi and H. gerlachei. While these two species agree in the narrow, uncleft, form of the endopod of the maxilla they differ in the following characters in addition to those mentioned by Kemp. Antennule : Stylocerite covers the otocyst in H. gerlachei. Mandible : In H. gerlachei the incisor process is larger in proportion to the molar part than in H . beaupresi and the molar part is intermediate between that of H. beaupresi, in which there are no teeth and a granular grinding surface is surrounded by stiff spines, and that of Periclimenes, having blunt teeth together with two short rows of spinules. Maxillipede 1 : Coxa and basis very distinct in H. gerlachei, scarcely distinguishable in H. beaupresi. Basal expansion of exopod very much less wide in H. gerlachei. Maxillipede 2 : Carpus without the sharp spine-like process seen in H. beaupresi, and with dactyl much narrower, and spines on it differently arranged. Maxillipede 3 : Arthrobranch vestigial in H. gerlachei. While it seems clear that P. diversipes should be transferred to the same genus as H. beaupresi, the separation of the latter from H. gerlachei is not so fully justified. Borra- daile distinguished, on insufficient grounds, Harpilius with type H. lutescens and including H. gerlachei, from Harpiliopsis, Borr. , with type H. beaupresi, and Tattersall, while rejecting Borradaile’s genera, suggested that it, might be necessary to form a new genus for H. gerlachei. With this Kemp did not agree. Having only examined two of the species of Harpilius I am not in a position to make any definite proposition on this point, but it is my opinion that the facts strongly support Tattersall’s suggestion. In the later stages of the larva the structure of legs 4 and 5 is the most striking feature of the two species of known parentage, and exactly the same form is found in the species described here under the designations R.S. I, B.R. II, B.B. Ill, D. Ill and D. IV. All these forms also agree in the structure of the maxilla, but differ among themselves in THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 29 details which may be regarded as specific. The following summary may now be given of the characters of Ancylocaris (if P. diversifies is excluded) : (1) Rostrum small, slender, with few or no dorsal teeth, and without median crest. (2) Carapace without hepatic spine and usually without branchiostegal. (3) Abdomen without dorsal spines. Anal spine absent. (4) Antennule without ventral spine (?). (5) Antennal scale, in Stage I. with or without inner papilla. (6) Maxilla, endopod with basal lobe ; exopod with 4 setae only in Stage I. outer margin bare in later stages. (7) Maxillipede 1. basis protuberant, without spines. (8) Maxillipede 3. in Stage I. with two long equal spines at end of dactyl. (9) Legs 4 and 5 very much longer than leg 3, subequal. Leg 4 without oxopod. propod with long terminal spines. Leg 5, propod without long terminal spines ; dactyl without inner basal spine. It is, of course, impossible to say if these characters are distinctive of the subgenus Ancylocaris only. The species described here as B.R. IV seems, from its general form of body and most of the appendages, to belong to Periclimenes , but it differs strikingly in the form of the rostrum, and particularly in the very unusual feature of having leg 5 very much smaller than leg 4. It must certainly represent a genus or subgenus distinct from Ancylocaris. Group II : Mesocaris. While there is no difficulty in accepting the larvae described above as a Periclimenes group closely related to each other, the remaining species are very difficult to deal with. There is here no basis of established fact, and it is impossible to determine how far the differences between the larvae are generic or specific. There is one striking feature which characterizes the larval genus Mesocaris (see Gurney, 1936, p. 621), namely the acute double flexure of the depressed body and abdomen, and the same flexure is seen in the three forms B.R. V, R.S. II and R.S. IV. They also agree in the general structure of legs 3-5, which are all of about the same length, and in having no basal lobe on the endopod of the maxilla ; but there is no other character common to all. None the less I include them, as a matter of convenience, in a single group. One of these forms (R.S. IV) is remarkable in having no spine on the antennal scale and, if this is also absent in the adult, it can only be referred to Onychocaris aualitica, Nobili, which it also resembles in the reduction of the rostrum. The two forms R.S. Ill and B.R. VI diverge greatly from the others. The former has not the flexure of the body characteristic of the group, and the enlargement of the abdominal pleura gives it a very distinct appearance. It is also unlike any other larva in having strong hooked spines on the propod of legs 3-4 instead of the usual straight slender spines. In Coralliocaris and other genera the dactyl is curved and provided with a hooked basal claw, but I cannot find any species described as having strong propodal claws. The peculiar modification of these legs does not afford, as one would expect it might, any clue to the genus. 30 GREAT BARRIER REEF EXPEDITION The second larva, B.R. VI, has the body flexed, and the legs of the form usual in Mesocaris, but it is remarkable for the great length of the rostrum, of the branchiostegal spine, and of the pleural spines of abdominal somites 1 and 2. Here, again, there is not sufficient ground for identification, but it may be pointed out that there is some general similarity to the strange genus Coutierea, Nobili. The single species of this genus, C. agassizi (Coutiere), was taken at Barbadoes, but there is no reason why the genus should not also occur in the Indo-Pacific. The great length of its rostrum, and antennal (or branchiostegal ?) spines, and the pointed abdominal pleura are striking features of the species. The eyes are overhung by a broadening of the rostrum which is regarded as produced by the connate supra-orbital spines. In the larva the supra-orbital spines are very small, and it may be that they are really lost in the adult and that the covering over the eye is simply an extension of the widening of the rostrum already present in the larva. Mesocaris, Ortmann. Gurney, 1936, p. 621, pis. iii, iv. Body very much flattened and flexed. Rostrum in late stages with high toothed crest. Carapace with small supra-orbital and with or without antennal and branchiostegal spines, hepatic spine absent. Telson with lateral spines ; apical spines 3 +3 (?). Anal spine absent. Antennule with ventral spine. Antennal scale in stage I with inner papilla and with outer proximal seta reduced. Maxilla, endopod with apical seta, without basal lobe ; exopod with five setae in stage I, with setae on outer margin in late stages. Maxillipede 1 in late stages with basis large, not protuberant, without spines ; endopod stout, with three distinct segments. Maxillipede 2, basis with two spines, the distal one strong ; apical spine of dactyl stout, with conspicuous feathering. Maxillipede 3 with single strong apical spine in all stages. Leg 4 with exopod ; propod with long terminal spines. Leg 5 not markedly longer than leg 4, propod with one long terminal spine ; dactyl with or without basal spine. Palaemonid B.R. V. (Text-figs. 124-133.) Locality. Barrier Reef Station 38. Description. Stage V ? Length 5-6 mm. Body rather depressed, and sharply bent at thorax and abdominal somite 3. Rostrum as long as segment 1 of antennule, with' one dorsal serrated tooth. Carapace with serrated supra-orbital spine and antennal spine. Pleura of abdominal somites 5 and 6 rounded. Anal spine absent. Telson four times as long as wide, with one pair of lateral and 4 + 4 terminal spines ; spines 1 and 3 large. Antennule, segment 1 with incipient stylocerite and ventral spine ; flagella small, unsegmented. Antennal scale narrow, with apical spine ; flagellum a little longer than scale. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 31 Fig. 124. — Stage V ?, lateral. Fig. 125. — ,, telson. Fig. 126. — ,, maxillipede 1. Fig. 127. — ,, maxilla. Fig. 128. — ,, mandible. Fig. 129. — Stage V ?, maxillule. Fig. 130. — ,, leg 5. Fig. 131.— „ „ 4. Fig. 132.— „ „ 3. Fig. 133. — ,, maxillipede 3. Maxillule, endopod with one small seta. Maxilla, endopod curved, with one long terminal seta, and no basal lobe or seta ; exopod without setae on outer margin. Maxillipede 1, basis without spines; endopod unsegmented. Maxillipede 2 and 3, dactyl with one strong apical spine ; exopods with four setae only. Legs 1 and 2 subchelate. Leg 4 with exopod, a little shorter than leg 5. Legs 3 and 4 with long spines at end of propod. Leg 5 not greatly elongated, propod without long spines ; dactyl with basal spine. Pleopods rather large, without setae. Palaemonid R.S. II. (Text-figs. 134-138.) Locality. Two specimens from plankton at Ghardaqa. Description. Length 4-55 mm. Last stage. Body sharply flexed at thorax and abdominal segment 3. Rostrum shorter than eye, with three dorsal teeth. Carapace with supra-orbital and branchiostegal spines. Pleura of abdominal somite 5 rounded, somite 6 with lateral spines. Anal spine absent. Telson nearly four times as long as wide, with two pairs of lateral spines and 5 + 5 terminal, of which spine 2 is very long. One of the specimens is about to moult and shows that the two median pairs of spines of the adult arise from spines 2 and 5 of the larva (Text-fig. 133). Antennule with small stylocerite and ventral spine ; flagella small, unsegmented. Antennal scale with terminal spine ; flagellum as long as scale. Maxillule, endopod with one very small seta. Maxilla, endopod slender, with long apical seta, but no basal lobe or seta ; exopod without setae on outer margin. Maxillipede 1, basis not very protuberant, without spines ; endopod unsegmented ; epipod small, not bilobed. Maxillipedes 2 and 3, dactyl with one strong apical spine. 32 GREAT BARRIER REEF EXPEDITION Text-figs. 134-138. — Palaemonid R.S. II. Fig. 134.- — Last stage, lateral. Fig. 137. — Last stage, palp of maxillule. Fig. 135. — ,, maxilla. Fig. 138. — ,, telson. Fig. 136. — ,, maxillipede 1. Legs 1 and 2 chelate, equal ; exopod of leg 1 with six setae. Leg 4 a little shorter than 5, without exopod ; spines on propod not very large. Leg 5 without long propodal spines ; dactyl with basal spine. Pleopods large, without setae. Remarks. In life this form is even more flexed than is shown in the figure, the thorax and abdomen forming an acute angle and the thorax being parallel to the last three somites of the abdomen. It swims back downwards and tail first. Colour : General colour to eye whitish yellow, the chromatophores dark olive-brown under microscope, but light yellow in reflected light. Palaemonid R.S. III. (Text-figs. 139-145.) Locality. One specimen from plankton at Ghardaqa. Description. Stage IV ? Length 4-65 mm. Body broad and depressed, but not strikingly bent. Rostrum much shorter than peduncle of antennule, with three dorsal teeth. Carapace with supra-orbital, branchio- stegal and pterygostomial spines. Abdominal somites 2 and 3 with pleura pointed and bent outwards so that they are visible from above ; somite 3 not swollen dorsally ; pleura of somites 4 and 5 rounded ; somite 6 with lateral spines. Anal spine absent. Telson parallel-sided, with two pairs of lateral spines and 4 + 4 apical, of which the outer three pairs are subequal. Antennule, peduncle curved, with large stylocerite and ventral spine ; flagella small, ansegmented. Antennal scale broad, with small distal spine ; flagellum nearly twice as long as scale. Maxillule with laciniae widely separated ; endopod with one small seta. Maxilla, endites reduced, with few setae ; endopod with long apical seta and no basal lobe or seta ; exopod very large, without setae on outer margin. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 33 Text-figs. 139-145. — Palaemonid R..S. III. Fig. 139. — Stage IV ?, dorsal. Fig. 143. — Stage IV ?, leg 1. Fig. 140. — ,, maxillule. Fig. 144. — ,, ,, 4. Fig. 141. — ,, maxillipede 1. Fig. 145. — ,, maxilla. Fig. 142. — „ leg 2. Maxillipede 1 , basis not protuberant, with one small spine ; endopod unsegmented ; exopod with five terminal setae. Maxillipedes 2 and 3 with one large terminal spine ; exopods of these and of legs 1-3 with six setae, the proximal pair unequal. Leg 1 with propod slightly produced ; leg 2 without trace of chela, the propod with two stout curved spines at base of dactyl. Legs 3-5 nearly equal, but leg 5 slightly shorter than 4. Legs 3 and 4 have each two strong propodal spines, as in leg 2, while leg 5 has one. Leg 4 with exopod. Pleopods present as small buds. Palaemonid R.S. IV ( Onychocaris ?). (Text-figs. 146-149.) Locality. One specimen from plankton at Ghardaqa. Description. Stage IV ? Length 3-5 mm. Body broad, depressed, much bent at thorax and somite 3. Rostrum very small, triangular in dorsal view, with elevated median crest, without teeth. Carapace with supra-orbital spines reduced to small blunt prominences ; branchiostegal and pterygo- stomial spines absent. Pleura of somites 5 and 6 rounded. Anal spine absent. Telson times as long as wide, without lateral spines ; end straight, with 4 + 4 spines, of which the outer pair is large and spines 2 and 4 vestigial, vi. 1. 3 34 GREAT BARRIER REEF EXPEDITION Antennule, peduncle broad at base, with large stylocerite and very small ventral spine. Antennal scale very broad, without spine ; flagellum about \\ times length of scale. Maxillule, endites rather wide apart ; endopod small, without setae. Maxilla, endopod slender, without basal lobe or seta ; exopod without setae on outer margin, narrow proximally. Maxillipede 1, basis slightly protuberant, without spines ; endopod two-segmented, with small seta at the joint. Maxillipedes 2 and 3 with single strong terminal spine, that of maxillipede 2 straight, with conspicuous feathering (Text-fig. 147). Text-pigs. 146-149. — Palaemonid R.S. IY. Fig. 146. — Stage IY ?, dorsal. Fig. 148. — Stage IY ?, maxillipede 2. Fig. 147. — ,, maxilla. Fig. 149. — ,, part of telson. Legs all equally long. Legs 1 and 2 not chelate. Leg 4 without exopod, with two rather stout spines at end of propod. Leg 5 dactyl without basal spine. Pleopods present, small. Colour : General colour greenish-yellow. Palaemonid B.R. VI ( Coutierea ?). (Text-figs. 150-154.) Locality. Barrier Reef Station 13. Description. Last stage. Length : rostrum, 2-5 mm. ; rest of body, 5-15 mm. Rostrum with four dorsal teeth, distal part bare. Carapace with small supra-orbital spine and enormous branchiostegal spine. Abdominal somite 1 with papilla on either side ; somites 2 and 3 produced into extremely long spines ; somite 3 very broad, THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 35 depressed dorsally, with pair of small lateral spines ; somites 4 and 5 with small pleural spines ; somite 6 long and slender, without lateral spmes. Anal spine absent. Telson four times as long as wide, without lateral spines ; apex with a large outer pair of spines and four small feathered spines between. Antenmile with large stylocerite and small ventral spine ; outer flagellum slightly cleft. Antennal scale narrow, with very long apical spine ; flagellum about as long as scale. Mandible with five movable spines. Maxillule, endopod small, curved, with one seta. Maxilla, endopod with long apical seta but no basal lobe or seta ; exopod with outer margin fringed with short setae. Fig. 150. — Last stage, lateral. Fig. 153. — Last stage, maxilla. Fig. 151. — ,, dorsal. Fig. 154. — ,, mandible. Fig. 152. — ,, telson. Maxillipede 1, coxa reduced; basis large, without spines; endopod with three distinct segments ; exopod with two outer basal setae, four apical setae and one lateral small seta ; epipod large, bilobed. Maxillipede 3, dactyl with single large spine. Legs 1 and 2 chelate, leg 2 the larger. Leg 4 with exopod. Leg 5 a little longer than leg 4, without propodal spines ; dactyl with basal spine. Pleopods large, without setae. Group III : Cryptoleander. For the three forms now to be described I have coined a name which is not intended as a generic designation but simply as a convenient term for reference. There is a certain general resemblance of these larvae to those of Leander and Retrocar is, the large serrated dorsal spines and the stout upturned rostrum giving them a marked Palaemonid appearance. As in Retrocaris the antennular flagellum becomes cleft in the later stages, and leg 5 is 36 GREAT BARRIER REEF EXPEDITION very long in comparison with leg 4. They differ from the Palaemonid group in the arrangement of propodal spines on leg 5 and in the absence of a basal lobe on the endopod of the maxilla. The branchiostegal spine is also absent. From the Periclimenes group they differ in almost every respect, and they also seem to have no near relation to the Mesocaris group, so that they are probably not members of the Pontoniinae at all. The resemblance to Anchistioides in general form is very striking ; but there are fundamental differences in the mouth-parts and structure of leg 5 which make it out of the question to include them in that genus. At the same time the resemblance in form seems to point to some degree of affinity and to suggest that Anchistioides may be more nearly related to the Palaemoninae than to the Pontoniinae. If these three forms cannot be included among the known genera of Palaemoninae or Pontoniinae, and do not belong to Anchistioides, what remains ? The genus or genera to which they belong must be common on the Barrier Reef and exist also in the Red Sea, but it does not necessarily follow that the adults have yet been seen. In a coral-reef region the difficulty in collecting is very great, and there may be many new genera yet to be discovered. As an instance of such difficulty may be mentioned the fact that though Anchistioides antiguensis is so commonly seen swimming at night at Bermuda, its real habitat is unknown and it has never been taken there in the daytime. Similarly the larva of Jaxea nocturna is not very uncommon at Plymouth but the adult has only once been captured there. On the other hand it is not impossible that the limit of the family Palaemonidae has been unduly restricted, and that it should be enlarged to include the Gfnathophyllidae . The three genera of the family have some remarkable characters special to themselves, and they also have a simple mandible without incisor process. It is the latter fact which has led to the association of Gnathophyllidae, Processidae and Crangonidae into the superfamily Crangonoida, but, if any weight can be given to larval characters then Processa is certainly not nearly related to Crangon, and the character of the mandible is not decisive. There is, in fact, a tendency to a reduction of the incisor process in some Palaemonidae and Alpheidae. Borradaile (1921) in describing Paratypton siebenrocki, Balss, points out certain features in which it resembles Gnathophyllum, and concludes “ I am inclined to place Paratypton near the point to which the Palaemonidae, Anchis- tioididae and Gnathophyllidae converge His figure of the mandible seems to indicate that the part regarded as molar is really compounded of the very minute reduced incisor and molar parts, and that the long “ incisor part ” is, as he himself suggests, a structure sui generis. Apart from the form of the mandible there seems to be less reason to associate Gnathophyllidae with Crangonidae than with Palaemonidae. It seems not unlikely that the Gnathophyllidae will be found to have larvae of Palaemonid type, and that they may 'even be among those described here. Palaemonid B.R. VII. (Text-figs. 155-162.) Locality. Barrier Reef Station 14 ; 6 specimens. Station 46 ; 1 specimen. Description. Stage V ? Length 8-4 mm. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 37 Rostrum longer than antennule, tinned slightly upwards, with serrated sphie below about middle, and another dorsal at base. Carapace with one large dorsal serrated spine ; a large supra-orbital spine but no antenanl or branchiostegal ; posterior angle produced into a large spine. A lateral ridge extends from the orbit nearly to the posterior angle. Text-figs. 155-162. — Palaemonid B.R. VII. Fig. 155. — Stage V ?, lateral. Fig. 156. — „ mandible. Fig. 157. — ,, maxilla. Fig. 158. — „ maxillipede 1. Fig. 159. — Stage V ?, tip of telson. Fig. 160. — „ part of leg 5. Fig. 161. — Stage IV ? (Ghardaqa), lateral. Fig. 162. — ,, dorsal. Abdominal somite 3 somewhat swollen dorsally, with a pair of large lateral spines curving downwards ; somite 5 with large curved lateral spines ; somite 6 without lateral spines. Anal spine absent. Telson 4| times as long as wide, without lateral or terminal spines, but simply bifurcated at end. Antennule with stylocerite and small ventral spine ; outer flagellum deeply cleft. Antennal scale without spine ; flagellum about If times as long as scale. Maxillule, endopod small, with two small setae. Maxilla, endopod small, with very small apical seta and no basal lobe or seta ; exopod with outer margin frmged with setae. Maxillipede I, basis very protuberant, without spines ; endopod of two segments, 38 GREAT BARRIER REEF EXPEDITION segment 1 with long seta ; exopod without basal setae, and five apical. Maxi 11 ipe.de. 3, dactyl with single large apical spine ; exopod with eight setae. Legs 1 and 2 chelate, nearly equal. Leg 4 smaller than leg 3, without exopod, and without long propodal spines. Leg 5 reaching beyond eye ; propod with two very long distal spines ; dactyl with basal spine. Lengths of segments in leg 5 : ischiomerus, 1-7 mm. ; carpus, 1-0 mm. ; propod, 1-6 mm. Pleopods small. Remarks. The Barrier Reef material contains also two other forms which, while agreeing in nearly all respects with the description given above, differ as follows : Form B. : Stations 55, 57. 3 specimens. Abdominal somite 3 produced dorsally into a large recurved spine. Form C. : Stations 33, 53. 2 specimens. Abdominal somite 3 without dorsal or lateral spines. One of these two specimens measures 10-6 mm. and is rather more advanced, having very large pleopods. The carpus of leg 5 is rather longer in proportion to the propod and ischiomerus. Some specimens from a plankton sample taken at Ghardaqa in July, 1935 (Text-figs. 156, 157), are younger than those from the Barrier Reef, measuring about 5 mm. Legs 3 and 4 are quite rudimentary, the rudiment of leg 3 biramous, leg 5 fully developed ; there are no pleopods. The telson is parallel- sided, deeply cleft at the end, the arms of the fork narrow and blunt-ended. Within the fork are two pairs of long and a pair of very short setae. Such a form of telson is very unusual and might well develop into the simple cleft form without setae found in the older specimens from the Barrier Reef. When first seen these Red Sea specimens were thought to belong to Anchistioides, but dissection of the mouth-parts makes such an identification improbable (see p. 43). I regard them as belonging to the same species as that described above, or at least to one very closely allied to it. Palaemonid B.R. VIII. (Text-figs. 163-169.) Locality. Barrier Reef Stations 33, 38, 50, 59. 14 specimens. Description. Stage VI ? Length 6-8 mm. Body rather bent at thorax and somite 3. Rostrum straight, much shorter than peduncle of antennule, serrated at end, and with one dorsal serrated spine. Carapace with two dorsal spines, the anterior one serrated ; supra-orbital and pterygostomial spines present. Abdominal somite 3 projecting over somite 4 ; somites 5 and 6 with lateral spines. Anal spine absent. Telson nearly four times as long as wide, with one pair of lateral spines ; outer apical spines large, forming a fork within which are two pairs of small feathered spines. Antennule with small stylocerite and very small ventral spine ; outer flagellum deeply cleft, shorter than inner. Antennal scale with long apical spine ; flagellum nearly four times as long as scale. Maxillule, endopod slender, with two apical setae. Maxilla, endopod without setae or basal lobe ; exopod with outer margin fringed with setae. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 39 Maxillipede 1 . coxa without setae ; basis very protuberant, without spines ; endopod with large basal segment distinct, bearing a long seta, distal part obscurely two-seg- mented ; exopod with one outer basal seta. Maxillipede 2, basis rather protuberant, with two long slender spines ; epipod present, small. Maxillipede 3, dactyl with one strong apical spine. Legs 1 and 2 chelate, leg 2 the larger. Exopods of maxillipede 3 and leg 2 with twelve setae, leg 1 with fourteen. In an older stage than that figured the chela of leg 2 is very long and slender, the dactyl twice as long as the palm. Leg 4 without exopod, slender, shorter than leg 3 and without spines on propod. Leg 5 extremely long ; coxa with anterior rounded knob : ischiomerus and propod about equal and about four times as long as carpus ; propod with strong distal spines ; dactyl with basal spine. Pleopods small. Text-figs. 163-169. — Palaemonid B.R. VIII. Fig. 163. — Stage VI ?, lateral. Fig. 167. — Stage VI ?, maxilla. Fig. 164. — ,, maxillule, palp. Fig. 168. — ,, leg 5, dactyl. Fig. 165. — ,, telson. Fig. 169. — ,, mandible. Fig. 166. — ,, maxillipede 1. The youngest specimen is only 3-8 mm., but has exactly the same form and spines on rostrum and carapace as the oldest. The uropods are fully developed, but there are no pleopods. While leg 5 is as large in proportion as in later stages, legs 3 and 4 are quite rudimentary. This stage no doubt represents Stage IV, and there is probably one moult between it and the stage figured, which is therefore regarded as Stage VI. An older stage has the pleopods very large, and is no doubt the last stage, so that there seem to be seven stages in all. Palaemonid B.R. IX. (Text-figs. 170-177.) Locality. Barrier Reef Stations 2, 9, 38, 44, 53. Description. Stage V ? Length 8 mm. 40 GREAT BARRIER REEF EXPEDITION Rostrum curving upwards, with one large dorsal spine at base. Carapace with large serrated dorsal spine ; supraorbital and pterygostomial spines present. Abdominal somite 3 not swollen dorsally ; somite 5 with lateral spines ; somite 6 without spines. Anal spine absent. Telson nearly four times as long as wide, without lateral spines, the outer terminal spines very large and forming a fork within which are six small spines. Antennule with very small stylocerite and large ventral spine ; outer flagellum slightly cleft, much longer than inner. Antennal scale with strong terminal spine ; flagellum about 1| times as long as scale. Fig. 170. — Stage Y ?, lateral. Fig. 171. — Fig. 172.— Fig. 173.— Fig. 174.— Fig. 175.— Fig. 176.— Fig. 177.— maxillipede 1. maxilla. mandible. end of antennule, showing cleft in outer flagellum, maxillule. leg 5, dactyl, a. From inside, b. From outside, end of telson. Maxillule, endopod slender, with two terminal setae. Maxilla, endopod slender, with long terminal seta but no basal lobe ; outer margin of exopod fringed with setae. Maxillipede 1, coxa with small seta ; basis very protuberant ; endopod of two segments, segment 1 with long seta. Legs 1 and 2 chelate, leg 2 the larger, its dactyl longer than palm. Exopods of maxillipede 3 and legs 1 and 2 with sik setae. Leg 4 about equal to leg 3, without exopod. Leg 5 very elongated ; coxa without inner papilla ; ischiomerus and propod equal, about three times as long as carpus. Propod with strong terminal spines ; dactyl with basal spine. Pleopods small. A younger specimen of 4-9 mm. of same general form and with leg 5 fully developed. ' had leg 4 a small rudiment. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 41 lc. Axchistioidinae. Anchistioides. The larva of Anchistioides antiguensis (Schmitt) in Stages I and II (Gurney. 1936) has some remarkable features which make it strikingly distinct from all Palaemonid larvae hitherto described. Apart from the reduction of the mouth-parts, which also characterizes the adult, the posterior spine on the carapace, the very large branchiostegal spine, and the dorsal spine on abdominal somite 3 are features which were not known at the time when the larva was described. While the paper was going through the press I found in plankton from Ghardaqa a Palaemonid larva (described above as B.R. VII, p. 38) with large posterior spines on the carapace which I supposed was a later stage of Anchistioides : but examination of the mouth-parts showed that this was an error. That the later stages of Anchistioides are hi fact much more Palaemonid in appearance than the earliest stages is confirmed by the discovery in the Barrier Reef material of late larvae which can with some confidence be referred to this genus. Two species are represented. Anchistioides Species I. (Text-figs. 178-184.) Locality. Barrier Reef Stations 15, 35, 44. 3 specimens. Description. Last stage ? Length 13-5 mm. including rostrum 4 mm. Fig. 178. — Last stage ?, lateral. Fig. 182. — Last stage ?, maxilla. Fig. 179.— „ part of antennule. Fig. 183. — „ maxillule. Fig. 180. — ,, maxillipede 1. Fig. 184. — ■ „ end of telson. Fig. 181. — ,, mandible. 42 GREAT BARRIER REEF EXPEDITION Rostrum long and straight, without spines dorsally, but with five ventral spines of which the proximal one points forwards and the distal four backwards. Carapace with small dorsal spine and large supra-orbital and branchiostegal spines. A small projection seems to represent the antennal spine, and there is a small spine at the posterior angle. Abdominal somite 3 with stout pointed dorsal process ; pleura of somites 5 and 6 rounded. Anal spine absent. Telson tapering, 4^ times as long as its anterior width ; without lateral spines ; posterior margin with a pair of strong outer spines confluent with the telson and forming a fork within which are six small spines. Antennule widened at base, with small rounded process representing stylocerite, without ventral spine ; outer flagellum deeply cleft, the thin inner branch reaching end of rostrum (Text-fig. 179), thick sensory part unsegmented. Antennal scale with small distal spine ; flagellum very long. Mouth-parts reduced. Palp of maxillule small, with one curved terminal seta. Maxilla without endites, the endopod small, without terminal seta ; exopod large, the outer margin fringed with small setae and hairs. Maxillipede 1, coxa not distinct, basis not protuberant, with small setae ; endopod small, unsegmented ; exopod with four outer basal setae, and four terminal. Maxillipede 2 with small rudiment of epipod ; endopod of three segments. Maxillipede 3, dactyl with single apical spine ; exopod with 14 setae. Legs 1 and 2 chelate, leg 2 very much larger than leg 1, the dactyl longer than palm. Leg 4 with exopod ; endopod smaller than leg 3. Leg 5, basis with posterior spine ; ischiomerus enormously produced, with a series of six recurved inner spines ; carpus a little shorter than propod, which is slightly dilated at end ; dactyl and its spine fused and forming a curved claw apposed to a stout spine on the propod to form a sort of pincer. Lengths of segments : ischiomerus, 7 mm. ; carpus, 2-25 mm. ; propodus, 3-15 mm. Pleopods large, without setae. Anchistioides Species II. (Text-figs. 185-192.) Locality. Barrier Reef Stations 24, 26, 40, 44. 6 specimens. Description. Stage IV ? Length 5-45 mm. Rostrum straight, shorter than peduncle of antennule, with very small ventral tooth. Carapace with large supra-orbital spines, but no branchiostegal spines or spines at posterior angle. Abdomen without dorsal spine ; pleura rounded. Telson 3| times as long as wide, without lateral spines, and with six feathered terminal spines. Anal spine absent. Antennule, peduncle with very small ventral spine ; flagella small, outer one not cleft. Antennal scale without spine, flagella about as long as body. Mouth-parts reduced. Maxillule, proximal endite much reduced ; endopod small, with one seta. Maxilla without endites ; endopod reduced to a small knob ; exopod large, without setae on outer margin. Maxillipede 1 , coxa not distinct ; endopod with distal segment marked off. Legs 1 and 2 not chelate. Leg 4 absent. Leg 5 exceedingly large ; ischiomerus reaching beyond antennal scale, without spines on anterior margin ; carpus long and THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 43 slender : propod divided into two segments with a strong spine at inner distal angle of segment 1 ; segment 2 not dilated, bearing, as in species 1, a large claw closing on to a strong spine on inner margin of segment. Pleopods absent. Uropods fully developed, exopod without spine. Last stage. Length 11 mm., including rostrum 2 mm. General form of body exactly as in Stage IV. Rostrum with one strong ventral spine ; eleven dorsal spines seen under skin, but ventral spines not sufficiently distinct to count. Antennule, outer branch deeply cleft, with long thin distal flagellum If times as long as thick basal part, and as long as peduncle. Antennal scale without spine. Text-figs. 185-192. — Anchistioides Species II. Fig. 185. — Stage IV ?, lateral. Fig. 186. — ,, maxilla. Fig. 187. — ,, mandible. Fig. 188. — ,, maxillule. Fig. 189. — Stage IV ?, telson, end. Fig. 190. — ,, maxillipede 1 Fig. 191. — Stage V ?, legs 1 and 2. Fig. 192. — Stage IV ?, end of leg 5. Legs 1 and 2 chelate, leg 2 much larger than leg l ; dactyl shorter than palm ; exopods with 18 setae. Leg 4 the same size as leg 3, without strong spines on propod, with exopod. Leg 5, lengths of segments : ischiomerus, 6-4 mm. ; carpus, 5-15 mm. ; propodus, 1-65 mm. Pleopods large. An intermediate stage of 7-65 mm. has the pleopods present as small buds, and leg 4 a small rudiment. Remarks. One of the late stages here described agrees with the earlier stages of A. antiguensis in having a dorsal spine on the abdomen, and it also has the same peculiar form of the mouth-parts, so that it can hardly be doubted that it belongs to the same genus, though there are great differences in the form of the carapace. In both the species described hypertrophy of leg 5 has produced an appendage of fantastic size, and terminating in a claw of remarkable construction. Such hypertrophy of leg 5 is characteristic of the 44 GREAT BARRIER REEF EXPEDITION l’.daemomdae, and the larva lias 111 fact the general facies of the Palaeniomdae of tlie Leandt r and ( 'ryptoleander groups, though it differs radically in form of the mouth-parts. Apart from the fact that the mouth-parts are rather more reduced in the Pontoniinae if. m in the l’alaemoninae there is nothing to suggest that Anchistioides is more nearly related to the Pontoniinae. Miss Gordon has discussed the systematic position of the genus and concludes that it is a true Pontoniid, but I suggest that it would be more satisfactory to regard it as forming a distinct subfamily. Having regard to the great difference between the larvae of Aunjlvcaris and those described here in the Mesocaris group it seems possible that, , veu within the Pontoniinae, some separation of genera into subgroups is required; but naturally nothing can be suggested until more of these larvae can be attached to their genera. II. ALPHEIDAE. The development of the Alpheidae is fairly well known through the work of Herrick, S.irs, Coutiere, Webb and Lebour. There is a strong tendency in the genus Synalpheus to abbreviation of development, and this is also seen in some species of Alpheus, but many, and perhaps most, of the species of Alpheus hatch as a perfectly normal zoea and pass through as many as nine larval stages (Lebour, 1932). The structure of the larva in such eases is remarkably uniform. Development is known in species of Alpheus, Synal - ////- '/.s and Athanas, and examination of the eggs of other genera by Coutiere indicates i li.ii the larva resembles that of Alpheus, and that development is not abbreviated. Such unidentified Alpheid larvae as are known ( Anebocaris , Diaphoropus, Bate) do not differ to any very marked extent from those of A. ruber, for example, but Anebocaris ancylifer, < outiere. is remarkable for its size (16 mm.) and is one of those large oceanic larvae which 1 1 * i u v i e r and (outiere suggest continued their larval phase far beyond the normal period. Tin- uniformity of structure is rather remarkable in view of the diversity of form ■■V nn th<- nearly allied family of Palaemonidae, and it is desirable to add to our knowledge b . d--M -ription of more larvae of known parentage. The form described below as Alpheid I * 1 1 'hows that there is some diversity, but unfortunately this larva cannot be named. PI" • Ibid Sea is extremely rich in Alpheids, as Coutiere has shown, but I was not very -ie i '"dul at Gliardaqa in obtaining larvae. A number of species were found, but some ... i. not breeding or had only just begun to breed, and it seems that Alpheids are pecu- 1 ditiieidi. to deal with in the laboratory, being very apt to throw off their eggs without 1,1 h 1 1 1 v • < >f one small species which is abundant in the coral many females were kept, ■ of tie in for many days, but not one of them produced larvae. A Ip hens ventrosus, M. Edw. (Text-figs. 193-198.) Locality. I In i a large, dark red, species found very commonly in the branches of the coral i '|,i generally a male and a female together. The eggs are numerous and not large in proportion to the size ol the animal. In a female of 34 mm., with early o *87 mm. by -62 mm. For the most part the eggs were all in early : ' ! ;t C'" ’ of one female hatched out in the laboratory on 23rd March. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 45 Description. Stage I. Length 2-4 mm. Rostrum absent. Carapace and abdomen without spines ; telson triangular, not deeply indented, with 7 -f 7 spines. Eyes sessile, oval in dorsal view. Antennule unsegmented. Antennal scale with four distal segments and ten inner and apical setae ; endopod less than half length of scale, with apical seta and small spine ; basis with small spine. Mandible rudimentary. Maxillule very small, lacinia 2 with two large spines. Maxilla with endites very reduced and endopod not marked off from stem ; exopod small, with five setae. Maxillipede 1. endopod unsegmented, basis large, protuberant, with four small spines. Maxillipedes 2 and 3, endopods unsegmented, without setae. Exopods with six setae ; in Text-figs. 193-198. — Alpheus ventrosus. Fig. 193. — Stage I, lateral. Fig. 196. — Stage I, maxillule. Fig. 194. — ,, telson. Fig. 197. — ,, maxilla. Fig. 195. — ,, antenna. Fig. 198. — ,, maxillipede 1. maxillipede 2 the proximal seta on the posterior side is reduced to a small hair. Behind the maxillipedes are two large leg rudiments, representing a biramous leg 1 and leg 5. Colour : All the specimens were almost colourless. Only four small red chromato- phores could be seen, but there was some diffuse yellow in the stomach region. Alpheus audouini, Coutiere. (Text-figs. 199-203.) Locality. A. audouini is very common on the tidal reef-flat at Ghardaqa, where it lives under stones and in crevices in the coral rock. The snapping of its claws can be heard every- where, but it is not easy to catch, since the rock is honey-combed with holes into which it disappears. It must also be very common in burrows in sand in deeper water since it was taken in some numbers when trawling over the Halophila bed at night, though never caught there by day. Description. Stage II. Length 2-75 mm. Carapace with small pointed rostrum and pterygostomial spine. Telson triangular, posterior margin straight, with 8 + 8 spines. 46 GREAT BARRIER REEF EXPEDITION Eyes stalked. Antennule with segment 3 of peduncle distinct ; exopod with one large and four slender aesthetes and a delicate hair, but no feathered seta. Antennal scale with four distal segments and 10 inner and apical setae ; a small knob on inner edge between setae and base ; endopod less than half length of scale, with apical seta and small spine. Mandible small, but with teeth. Maxillule very small, with two large spines on segment 2. Maxilla with three well-marked endites bearing small setae ; endopod with vestige of basal lobe ; exopod small, with five setae. Maxillipede 1, endopod very small, with apical seta. Maxillipede 2, endopod of four segments, with apical spine. Maxillipede 3, endopod very long and slender, with long apical spine ; exopods with 4.6.6 setae. Leg 1 a large biramous rudiment without setae. Leg 2 a small biramous rudiment ; leg 3 a small papilla ; leg 4 absent. Leg 5 a large, blunt-ended rod. Uropods traceable within telson. Text-figs. 199-203. — Alpheus audouini. Fig. 199. — Stage II. Fig. 202. — -Stage II, maxillule. Fig. 200. — ,, telson. Fig. 203. — ,, maxilla. Fig. 201. — ,, antenna. Remarks. Among large numbers of specimens from two broods nearly all are as described above. Two only were found partially moulted from a normal Stage I, having sessile eyes and telson with 7+7 spines. All breeding females were examined every morning and evening, and it is clear that in this case there is a first stage which moults within a few hours into Stage II. Brooks and Herrick (1892, p. 361) noted the very short duration of Stage I in A. minor (=A. bermudensis, Bate) and it may be well that this is usual in the Alpheidae. Miss Lebour (1932) found that the first larva in Athanas nitescens and Alpheus ruber had the rostrum and stalked eyes characteristic of Stage II, and it seems likely that here also Stage I has already been passed through. There must, however, be some variation in this respect since Miss Webb (1921) found a normal Stage I of A. nitescens in plankton. I have described (1927) three stages, including a normal Stage I of this species from the Suez Canal. Though the identification was then only a matter of guesswork I have no doubt now that it was correct. Development probably proceeds as in A. ruber through a number of stages, but no later larvae which could be attributed to it were seen at Ghardaqa. Larvae of another species, distinguishable by the presence of a supra-orbital spine, were common in plankton from outside the reefs, but not seen in the lagoon. I attribute these with much doubt to A. pacificus, Dana. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 47 A. pacificus, Dana ? (Text-figs. 204-214.) Coutiere, 1905, p. 909. Description. Larva in last stage. Length 6-3 mm. Carapace with rostrum shorter than eves and with snpra-orbital and pterygostomial spines. Abdominal pleura rounded ; somite 6 more than twice as long as deep. Telson more than three tunes as long as wide, without lateral spines, and with 4 + 4 apical spines, of which the outer pah' are very long. Anal spine absent. Fig. 204. — Last stage. Fig. 205. — ,, carapace from above. Fig. 206. — ,, end of telson. Fig. 207. — ,, maxillule. Fig. 208. — Post-larval I. Fig. 209. — ,, antennule. Fig. 210. — Post-larval I, telson. Fig. 211. — - ,, mandible. Fig. 212. — - ,, maxillule. Fig. 213. — ,, maxilla. Fig. 214. — - ,, maxillipede 1. Endopod of antenna longer than scale, with basal segment distinct. Mandible without palp. Maxillule, endopod small with large apical seta ; lacinia 2 with three strong spines. Maxilla, exopod with outer margin without setae ; endites with several setae , endopod with basal seta. Maxillipede 1, basis with four marginal spines. Legs 1 and 2 indistinctly segmented, with small chelae. Leg 5 reaching with its apical spine to beyond the eyes. Epipods present on legs 1-4. Pleopods large, without setae. Uropods, exopod with small outer apical spine. 48 GREAT BARRIER REEF EXPEDITION Colour : Bright yellow below eyes, in labrum and at bases of maxillipede 3 and leg 3. Two small red chromatophores at posterior end of carapace, and larger ones dorsally in abdominal somites 1^. From this stage moults were obtained to post-larval. Post-larval Stage I. Length 6 mm. Carapace with small rostrum, and partly, but not wholly, overhanging the eyes. Telson nearly three times as long as wide, with two pairs of dorsal spines and two pairs of terminal spines, between which there are three pairs of long feathered setae. Anal spine absent. Antennule, segment 2 longer than segment 3 (42 : 30) ; segment 1 with large inner spine and sharp-pointed stylocerite ; otocyst widely open above (Text-fig. 209). Mandible without palp. Mouth-parts and maxillipedes of adult form. Maxillipede 1 and 2 with epipod ; maxillipede 3 with rudiment of arthrobranch ; epipods on legs 1-4. The exopod of maxillipede 3 bears setae, and is not reduced as is so often the case in post-larval Stage I. Leg 1 on left with very large chela, the dactyl about two-thirds length of palm and broadening near end. Right chela slender, the dactyl nearly as long as palm. Leg 2 carpus of five segments of following lengths — -30, 19, 13, 12, 27. Legs 1-3 with spine on posterior margin of ischium ; dactyl simple and slender. Colour : Almost colourless, but with yellow in same places as in larva. Remarks. It is not to be expected that the specific characters of the adult would be developed at this stage sufficiently for correct identification, especially in so difficult a genus. The form of the chela is probably more or less like that of the adult, at least in its general proportions ; but the carpus of leg 2 certainly becomes more slender and the proportional lengths of the segments may also change Assuming, however, that these proportions may be relied upon, most of the species known from the Red Sea are excluded, since it is more usual to find segment 1 shorter than segment 2 throughout the genus. In A. ; pacificus , according to Coutiere, segments 1 and 2 have the proportions 1*27 : 1, whereas they are 1-58 : 1 in my specimen. The general form of the dactyl of the large chela is the same, though the propod is very much stouter in the adult A. jpacificus. Synalpheus triunguiculatus (Paulson). (Text-figs. 215-232.) Locality. This small, dark-red Alpheid was found in small numbers on the inner reefs at Ghardaqa in coral of the genus Stylophora. One female hatched eggs on 23rd March. A number of the larvae moulted two days later to Stage II, and one specimen in Stage II and another in Stage III or IV were taken in plankton. Coutiere (1899, p. 450) regards S. triunguiculatus as a variety of S. neptunus, Dana. Both are common at Djibouti, but S. neptunus is found generally in sponges, and has large eggs, with a very advanced egg-larva, while he found S. triunguiculatus living in Stylophora and having smaller eggs, with a zoea larva. He noted the large size of the pleura of somite 2 (p. 455). THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 49 Text-figs. 215-232. — Synalpheus triunguiculatus . Fig. 215.— Stage I. Fig. 216. — „ telson. Fig. 217. — ,, antenna. Fig. 218. — „ maxillule. Fig. 219. — ,, maxilla. Fig. 220. — „ maxillipede 1. Fig. 221. — Stage II. Fig. 222. — „ telson. Fig. 223. — ,, antennal scale. Fig. 224. — Stage II, mandible. Fig. 225. — ,, maxilla. Fig. 226. — ,, maxillipede 1. Fig. 227. — Stage IV ? Fig. 228. — ,, bead, dorsal. Fig. 229. — „ telson. Fig. 230. — ,, leg 1. Fig. 231.— „ „ 2. Fig. 232. — „ „ 3. vi. 1. 4 50 GREAT BARRIER REEF EXPEDITION Description. Stage I. Length 2-5 mm. Rostrum very broad, shorter than eyes, abruptly narrowed at end into a small point. Carapace without spines or marginal teeth. Abdominal somite 2 with conspicuously large pleura. Telson very narrow, nearly rectangular, setae 1 and 2 directed outwards and forwards, setae 3-5 very long ; seta 2 feathered on both sides. Eyes free. Antennule with peduncle segmented. Antennal scale unsegmented, with two outer and ten inner and terminal setae ; a very small tubercle on inner margin next to proximal seta ; endopod more than half as long as scale, with two short apical setae. Mandible rather rudimentary, but showing distinction of incisor and molar parts. Maxillule very small, but with spines on endites. Exopod of maxilla with four setae in front and one large proximal seta ; endopod with rather large basal lobe ; three endites with small spines. Maxillipede 1 much reduced ; exopod without setae ; basis with four small spines. Maxillipedes 2 and 3, endopod unsegmented, without setae ; exopods with 5 and 6 setae, four apical. Legs 1 and 2 rudimentary, the exopods large, without setae. Leg 5 very long, turned forwards, without apical spine. Stage II. Length 2-6 mm. Specimen from plankton 2-87 mm. Rostrum broad at base, narrowed in front into an upturned spine. Abdominal somite 2, pleura very large and partly covering somites 1 and 3. Pleura of somites 4 and 5 with ventral point. Telson jointed to somite 6, parallel-sided, more than twice as long as wide, without lateral spines and with only 7 + 7 apical setae, of which the innermost is minute. Anal spine absent. Antennule with very small endopod ; base not enlarged. Antennal scale with 13 inner and apical setae and only one outer seta ; endopod with two apical setae. Mouth-parts almost unchanged ; exopod of maxilla the same. Maxillipede 1, exopod without setae. Exopods of maxillipedes 2 and 3 with six setae ; endopods with setae and terminal spine. Legs 1 and 2 fully formed ; legs 3 and 4 rudimentary. Leg 5 very long, with long smooth apical spine. Uropods developed ; exopod not jointed to basis, with six setae ; endopod jointed, without setae. This stage is equivalent to the normal Stage III, and this appears to be another case of the omission of the ordinary Stage II. In view of my observation of the very short duration of Stage I in A. audouini it seems probable that there is a similar speeding-up of metamorphosis in S. triunguiculatus, but my attention was not directed to this point on the spot, and I have no evidence of this supposed first moult. Stage III (or IV ?). Length 5-0 mm. Rostrum and carapace unchanged. Abdominal somites 4 and 5 with sharply-pointed pleura. Telson very long and narrow, without lateral spines, and with 4 + 4 apical spines. Anal spine present. Antennule, endopod longer than exopod, with small apical segment. Endopod of antenna longer than scale, with proximal and distal segments marked off ; scale without apical spine. Mouth-part apparently unchanged. Legs 1 and 2 with large chelae ; legs 3 and 4 fully developed, with exopods. Leg 5 with long spine reaching beyond eyes. Pleopods present, pleopod 1 very small. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 51 This specimen, which is rather decayed and damaged, was about to moult to post- larval. It is so much further advanced in development than Stage II that there is probably an intermediate stage which was not found. Assuming also a missing Stage I, there would then be five stages in all, as in some Palaemonidae. Alpheid R.S. I. (Text-figs. 233-237.) Locality. Ghardaqa plankton. Description. Length 3-7 mm. Rostrum large and broad, longer than eyes, and overhanging them. Carapace and abdomen without spines ; pleura rounded. Telson nearly three times as long as wide, without lateral spines ; with 4 + 4 apical spines of which the outer pair are very long. Anal spine present. Text-figs. 233-237. — Alpheid R.S. T. Fig. 233. — Stage IV ? Fig. 236.— Stage IV ?, maxillipede 1. Fig. 234. — „ telson. Fig. 237. — ,, leg 3, dactyl. Fig. 235. — ,, maxilla. Antennule with minute rudiment of stylocerite, but not widened at base. Antennal scale with small apical spine ; endopod a little shorter than scale, without basal segment ; basis without spine. Mandible small but toothed. Maxilla with three well-developed endites ; endopod without basal seta ; exopod with four large setae anteriorly, outer margin smooth, narrowed behind and bearing here one small seta. Maxillipede 1, endopod small, unsegmented; basis with four very large marginal spines ; rudiment or epipod present. Maxillipedes 2 and 3 endopods normal, with four segments ; exopods with six setae. In the single specimen seen legs 1 and 2 are large and chelate on the right side, but without chelae on the left, though the propods are enlarged. Legs 3 and 4 alike, with short dactyl and apical spine (Text-fig. 237). Leg 4 with exopod without setae. Leg 5 very long and slender, reaching with its spine to end of eyes. Pleopods present, small. Uropods well developed, exopod with 12 setae, and endopod with 10. 4§ vi. 1. 52 GREAT BARRIER REEF EXPEDITION Colour : General colour olive-brown, but with red in the chromatophores on somite 4 and at base of leg 3. Antennule yellow. Remarks. While this is quite certainly an Alpheid larva, and very closely resembles the larva of Synalpheus, the absence of setae from the exopod of leg 4 is peculiar and the asymmetry of legs 1 and 2, if it is not merely abnormal, is so unusual a feature that I feel this larva may belong to another genus altogether. Alpheid R.S. II. (Text-figs. 238-241.) Locality. Ghardaqa plankton. Description. Stage II ? Length 4-45 mm.-4-9 mm. Text-figs. 238-241. — Alpheid R.S. II. Fig. 238. — Stage II, telson. Fig. 240. — Stage II, maxillule. Fig. 239. — ,, maxilla. Fig. 241. — Stage III. Carapace with small rostrum ; anterior angle pointed. Telson segmented from somite 6, about as wide as long, with 7 + 7 spines and in one case a minute median tubercle which may represent spine 8. Eyes stalked. Antennule, peduncle of three distinct segments, endopod as long as exopod. Antennal scale unsegmented, with small spine and many setae ; endopod longer than scale, stout, with basal segment distinct, and also a small distal segment. Mandible small, but with toothed incisor part. Maxillule, segment 2 with two large spines. Maxilla with endites reduced ; endopod with two setae ; exopod fringed with setae. Maxillipede 1, basis with four spines, epipod present. Maxillipedes 2 and 3 with very small apical claws. Legs 1 and 2 chelate, leg 1 much larger than 2. Legs 1-3 with exopods, without setae. Leg 5 not larger than 4. Stage III ? Length 4 -5-4 *8 mm. Carapace as before. Telson widest at end, with 6 + 6 spines. No anal spine. Mouth-parts and thoracic appendages unchanged except that chelae are larger and exopods on legs 1-3 have setae. In one specimen near moulting the right chela is very much the larger, and has a large knob on the inner edge of the dactyl. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 53 Pleopods large, with incipient setae. Uropods present, the endopod not jointed and without setae. Colour : General ground-colour greenish-yellow, with black dots and a few red chromatophores. Median ventral chromatophores hi abdomen red. One specimen of Stage III died in the moult to post- larval, but the new stage is not sufficiently free from the old skin to show its characters. It can, however, be seen that there is a rudimentary arthrobranch on maxillipede 3 and no epipods on the legs. Remarks. The general resemblance of this larva to that of “ Alpheus saulcyi var. brevicarpus ” described by Herrick is very close, and this, with the absence of epipods from the legs make it practically certain that it is the larva of a species of Synalpheus. It differs from Herrick’s species in having no exopod on leg 4, and from S. triunguiculatus in the same way, but also in general form and in having leg 5 not specially elongated. It is, in fact, a quite distinct type of larva, and it is unfortunate that it cannot be identified. Only the two stages described were seen, and moults were obtained of Stage II to Stage III. If there is an earlier stage it is, no doubt, a transitory one ; but, in any case development is much shortened since “ Stage III is certainly the last larval stage. There seems to be no growth between the stages, the larva probably living on yolk and not by capture of food. Alpheid B.R. I. (Text-figs. 242-248.) Locality. Barrier Reef Station 50. Description. Length 7*55 mm. Rostrum as long as eyes, very broad at base where it bears three large marginal Fig. 242. — Last stage. Fig. 246. — Last stage, maxilla. Fig. 243. — ,, dorsal view of head. Fig. 247. — ,, maxillipede 1 Fig. 244. — „ end of telson. Fig. 248. — leg 4, dactyl. Fig. 245. — ,, maxillule. 54 GREAT BARRIER REEE EXPEDITION teeth on either side. Carapace with pterygostomial spine. Abdominal pleura rounded. Telson four times as long as wide, without lateral spines and with 4 + 4 terminal spines, of which the outer pair are very long. Anal spine absent. Antennule with stylocerite, without ventral spine ; endopod longer than exopod. Antenna, endopod longer than exopod with basal segment distinct. Mandible small, but with teeth. Maxillule, lacinia 2 with three large spines. Maxilla, endopod with basal seta ; exopod with part of outer margin bare. Maxillipede 1, basis with four small spines ; epipod very large. Maxillipede 2, endopod of four segments, with strong apical spine ; epipod small. Legs 1 and 2 with chelae. Legs 1-4 with exopods ; epipods on legs 1-4. Leg 5 extending with spine beyond antennules. Pleopods large, without setae. Uropods fully developed. Remarks. One specimen of this Alpheid was found in a plankton sample from Station 50 of the Great Barrier Reef Expedition. While it is in other respects a perfectly normal larva of the Alpheus type it differs from all other known larvae in the strongly toothed margins of the rostral hood. Athanas djiboutensis, Coutiere. (Text-figs. 249-252.) Locality. The adult is found commonly in pools on the reef-flat at Ghardaqa, but only one was taken with eggs. From these rather large eggs about 15 larvae were hatched on 25th March. Description. Stage I. Length 1*5 mm. Text-figs. 249-252. — Athanas djiboutensis. Fig. 249. — Stage I. Fig. 251. — Stage I, maxillule. Fig. 250. — ,, antenna. Fig. 252. — ,, maxilla. THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 55 Rostrum absent ; carapace and abdomen -without spines. Telson rather narrow, slightly indented, with 7 + 7 setae, of which the fourth is the longest. Peduncle of antennule imsegmented. Antennal scale segmented, with two outer and ten iimer and terminal setae ; endopod a short spine-like process. Mandible rudimentary. Maxillule and maxilla very much reduced, apparently not functional, with few setae. Maxillipede 1, endopod minute, with one apical seta ; basis without setae or spines. Maxillipede 2 and 3, endopod rather small, of three segments, with terminal curved spine. Exopods with 4.6.6 setae. Exopod of maxillipede 2 with proximal outer seta very small. Leg 1 a large biramous rudiment. Leg 5 a large, blunt-ended rudiment. Legs 2-4 absent. Uropods visible mider skin. Colour : Conspicuously red in eyes, somite 3 and telson, but yellow bands under reflected light. The large dorsal chromatophores have both red and olive-brown branches. No ventral chromatophores in abdomen, but a large richly-branched one at base of maxillipede 3. Remarks. It will be noted that this is a quite normal Stage I larva very much like that of A. dimorphus (Gurney, 1927). In A. nitescens (Lebour, 1932) the first larva from the egg has a rostrum and stalked eyes, but the telson has 7 + 7 setae. It is suggested above that there may be, in that species, a Stage I which moults very soon after hatching into Stage II. Alpheid D. I. (Text-figs. 253-256.) Locality. Discovery Station 1580. 8° 44' S., 41° 50' E. Description. Length : rostrum 2-85 mm. ; rest of body 9-6 mm. Rostrum longer than antennules, perfectly straight, slightly widened at base, with a marginal tooth on either side (supra-orbital spine). Carapace with median dorsal tubercle and small pterygostomial spine. Abdominal somites with large pleura, each with a Fig. 253. — Last stage ? Fig. 255. — Last stage ?, maxilla. Fig. 254.— ,, part of telson. Fig. 256. — ,, maxillipede 1. 56 GREAT BARRIER REEF EXPEDITION large tooth, recurved on somites 1-4. Somite 6 very long, with posterior lateral triangular projection over base of telson. Telson very long and narrow (2 mm. x 0-3 mm. at base), without lateral spines, and with 4 + 4 apical spines ; outermost spine very large, inserted below a median projection of the telson which bears the remaining six spines. Anal spine absent. Antennule, peduncle segmented, with stylocerite and minute ventral tooth ; endopod longer than exopod, with five segments, the two basal very long ; exopod with small distal segment. Antennal scale narrow, with strong apical spine ; endopod longer than scale and with basal segment distinct. Mandible without palp, incisor process toothed. Maxillule well developed, lacinia 2 with four strong spines. Maxilla, exopod with outer margin bare ; inner lobes with spines ; endopod with small basal seta. Maxillipede 1 with large bilobed epipod ; basis with four strong spines ; endopod slender, unsegmented, with long setae ; exopod with six setae. Maxillipede 2, endopod of five segments, ischium very short ; rudiment of epipod present. Maxillipede 3, endopod of four segments ; exopod with 10 setae ; rudiment of arthrobranch present. Legs 1-4 with exopods ; leg 1 with large chela, right chela the larger. Leg 2 long and slender, with chela. Pleopods present. Exopod of uropods with small spine. Leg 5 lost, but coxa and basis large ; coxa with large posterior spine. Kemarks. This larva is of rather special interest since it so closely resembles Anebocaris ancylifer, Coutiere, in its large size, long rostrum and large toothed abdominal pleura. It differs in the form of these pleura, and in the very much shorter carapace and longer abdominal somite 6. For comparison I give a figure of a specimen of Coutiere’s species of about the same size as the specimen here described (Text-fig. 257). No suggestion can, of course, be made as to the parentage of these remarkable forms, but they do not differ in essentials from the larvae of Alpheus, and they may well be found to fall within the genus. Alpheid ? D. II. (Text-figs. 258-265.) Locality. “ Discovery ” Station 276 : 5° 54' S., 11° 19' E. „ „ 277 : 1° 44' S., 8° 38' E. „ „ 703 : 7° 17' N., 28° 02' W. „ „ 704 : 3° 37' N., 29° 14' W. „ „ 705 : 0° 03' N., 30° 36' W. 2 specimens. 1 specimen. 3 specimens. 25 32 55 55 THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 57 Description. Oldest stage,, length 16 mm. Carapace narrowing forwards in side view, with anterior and posterior dorsal papillae and supra-orbital spines. Rostrum straight, without teeth. Abdomen bent at somite 3 ; somite 3 somewhat protuberant dorsallv ; somite 6 very long, very much compressed laterally and almost knife-like. Telson deeply hollowed below, nearly four times as long as wide, with two pairs of lateral spines and 3 + 3 apical, the inner pair feathered. In a younger specimen there are 5 + 5 apical spines, of which 2 and 4 are the longest, and are the two long spines of the older stage. Anal spine present. Text-figs. 258-265. — Alpheid ? D. II. Fig. 258. — Larva of 16 mm. “ Discovery " Station 276. Fig. 262. — Larva of 16 mm., telson (part). Fig. 259. — ,, mandible. Fig. 263. — ,, dactyl of leg 5. Fig. 260. — - ,, maxillule (part). Fig. 264. — - ,, ,, ,, 3. Fig. 261. — antennule (part). Fig. 265. — ., endopod and endites of maxilla. Antennule with small rounded stylocerite ; outer branch deeply cleft. Antenna, flagellum more than twice as long as scale, which has a very small apical spine. Maxillule, endopod reduced to a rounded lobe bearing a minute seta. Maxilla, exopod very large, with setae on outer margin ; endopod small without basal lobe or seta ; three inner laciniae. Maxillipede 1 with large epipod ; basis long and straight ; endopod short, broad, unsegmented, or with faintly marked division into three segments ; exopod without outer basal setae. Maxillipede 2 with large flat epipod ; endopod of five segments ; exopod with six setae. Maxillipede 3 with epipod and rudimentary arthrobranch. There are no artkrobranchs on the legs, but epipods are present on legs 1-4. Legs 1 and 2 chelate, leg 2 the larger ; exopod of leg 1 with 24 setae. Legs 1-4 with exopods. Leg 5 much larger than leg 4, slender ; dactyl with basal spine. Pleopods large ; pleopod 1 , endopod very small, without appendix interna. 58 GREAT BARRIER REEF EXPEDITION Relation of Palaemonidae to Alpheidae. Coutiere (1899), whose opinion, based upon a profound comparative study of the Alpheidae, is entitled to the utmost respect, concluded that the least modified genera of the family are among the most primitive Caridea, and his comparison with other families led him to the conviction that they were most nearly related to the Hippolytidae. Borra- daile (1907) included the Palaemonidae and Alpheidae in his superfamily Palaemonoida, together with the Hippolytidae and Rhynchocinetidae, and at that time he included Hymenocera in the Palaemonidae, while Gnathophyllum fell within the Crangonoida. Balss (1927) accepted Borradaile’s system with the exception that he transferred Hymeno- cera to the Gnathophyllidae, which he retained in the Crangonoida. In most published systems of Caridea there seems to be a tendency to regard the Alpheidae as more nearly related to the Hippolytidae than to the Palaemonidae. The larva of Rhynchocinetes is now known, though not yet described, and has characters which completely exclude it from the Palaemonoida, so that, in considering the position of the Alpheidae it is only necessary to bring the Palaemonidae and Hippolytidae into comparison. Coutiere does not seem to have considered the Palaemonidae as possible near relatives of the Alpheidae although in many, indeed in most, cases the characters to which he draws attention as evidence for Hippolytid affinity would have equal weight if Palae- monidae were substituted for Hippolytidae. The only characters of importance in which the Alpheidae as a whole differ from the Palaemonidae are the presence of epipods on the legs and the segmentation of the carpus of leg 2. Segmentation of the carpus is found in a number of unrelated genera of Caridea and is not of itself a feature of first-rate systematic importance. The absence of epipods from the legs of all Palaemonidae and their presence in most Alpheidae is an important difference, but it loses some of its weight when one recalls that they are also absent in Synalpheus, Cheirothrix and Ogyrides (which Coutiere speaks of as “ in many respects a true Hippolytid ”). The resemblance in general form and in hypertrophy of the chelae of some Pontoniinae to the Alpheidae is very striking, and when one finds one of them ( Coralliocaris graminea) snapping its claws like an Alpheid, it is difficult to suppose that the resemblance can be entirely a matter of convergence to a similar habitat. On the basis of adult structure the phylogenetic grouping of the Caridea is exceedingly difficult, and if any evidence can be gathered from the larval phase it should be valuable, though it has not hitherto been taken very seriously. Attention has already been called (Gurney, 1924, p. 131) to the close resemblance between Palaemonid and Alpheid larvae in some important characters, and a more detailed comparison may now be made in the light of the richer material available. Apart from the genus Lysmata the precocious development of leg 5 and its excessive size are confined to the Alpheidae and Palaemonidae. Within these two families the propod is long and slender and the dactyl usually produced into a long spine, whereas in Lysmata the propod is dilated into a paddle-like organ and the dactyl is very much reduced. The great development of leg 4 seen in Periclimenes finds something of a parallel in a Lysmatid in which leg 4 has a paddle-like propod as well as leg 5 ; but the unlikeness in structure and function of these legs in the two types suggests that the disproportion in size and precocious development of leg 5 have been separately acquired. While typical THE LARVAE OF THE DECAPOD CRUSTACEA— GURNEY 59 Alpheid larvae have a very uniform type of styliform leg 5, there are genera of Alpheidae, just as there are genera of Pontoniinae, in which leg 5 is not larger than legs 3 and 4, and it seems likely that there may be two groups within the family of which Alpheus and Synalpheus would be typical respectively. The general reduction of the maxillae and maxillipede 1 which characterizes all the Palaemonidae, and is carried to excess in some Alpheidae, is not found in any other families except where development is abbreviated ( e.g . Pasiphaea). Invariably the endopods of the maxillule and maxilla are reduced and misegmented, and at least the proximal lacinia of the maxilla is absent. In maxillipede 1 the coxa is reduced or scarcely traceable, and the endopod is small and often unsegmented. In the Hippolytidae this reduction does not take place. The endopod of the maxillule may be unsegmented, but it always retains evidence of the loss of segmentation in the presence of one or two setae on its inner margin. The endopod of the maxilla, though it may not be segmented, has a series of three inner lobes bearing setae. The inner laciniae are also well developed, lacinia 2 being distinct, and each bears numerous setae. Lastly the endopod of maxillipede 1 is long, and generally, if not always, shows four segments. The antenna of Palaemonidae and Alpheidae in Stages I and II has always the exopod distinctly segmented, and there is often, in both families, a peculiar papilla on the inner margin. The endopod is a straight rod bearing a long seta and a small spine. In Hippo- lytidae the exopod may be unsegmented, or with slight traces of segmentation ( Hippolyte ), or it may be slender and segmented ( Caridion , Lysmata), but there is no inner papilla The endopod is not the same in the different genera, but it never has the combination of seta and spine seen in Palaemonidae and Alpheidae. In these two families the exopods in stage I always bear four apical setae symmetrically disposed in two pairs ; but in Hippolytidae one seta of the proximal pair is absent so that there are three apical setae, an arrangement which is also seen in Processa, Leptochela and some, if not all, Pandalidae. Those who have to discuss the systematics of adult Caridea are commonly forced to rely upon characters which seem to be of very small importance, for example, form of rostrum, shape of the epipod of maxillipede 1 , etc. The points of similarity and difference which I have given here are at least positive and unequivocal, and I submit that they suffice to prove that the Palaemonidae and Alpheidae are very closely related, and that they are not nearly related to the Hippolytidae. This conclusion is based upon consideration of larvae which, although not certainly identifiable, can with some confidence be assigend either to Palaemonidae or to Alpheidae, but the close relation between the two families is shown perhaps most strikingly by the larva described as Alpheid ? D. II. In this case we have a form which, while the salient characters are, on the whole, Palaemonid, has also something of the form of an Alpheid, and a series of epipods exclude it from the Palaemonidae. It is impossible to say to which of the two families it belongs, but it is quite out of the question to refer it to any other family of which the larvae are known. 60 GREAT BARRIER REEF EXPEDITION REFERENCES. Balss, H. 1927. Decapoda. In Kiikenthal, Handbuch der Zoologie, Bd. III. Borradaile, L. A. 1907. On the classification of the Decapod Crustacea. Ann. Mag. Nat. Hist. (7) XIX, pp. 457-486. 1917. On the Pontoniinae. Trans. Linn. Soc. XVII, pp. 323-396. 1921. On the Coral-gall Prawn Paratypton. Mem. Lit. Phil. Soc. Manchester, LXY, no. 11 Brooks, W. K., and Herrick, F. H. 1891. The embryology and metamorphosis of the Macrura. Mem. Acad. Sci. Wash. Y, pp. 321-576. Coutiere, H. 1899. Les Alpheidae. Ann. Sci. Nat. Zool. (8) IX, pp. 560. 1905. Les Alpheidae. Fauna and Geog. of Maldives and Laccadives, II, pp. 852-921. 1907. Sur quelques formes larvaires enigmatiques d’Eucyphotes. Bull. Mus. Oceano. Monaco, no. 104, pp. 70. Gordon, I. 1935. On new or imperfectly known species of Crustacea Macrura. J. Linn. Soc. XXXIX, pp. 307-351. 1936. On the Macruran genus Rhynchocinetes. Proc. Zool. Soc. 1936, pp. 75-88. Gurney, R. 1924. Decapod Larvae. British Antarctic (“ Terra Nova ”) Expedition, 1916, Zoology, VIII, no. 2, pp. 37-202. 1927. Larvae of the Crustacea Decapoda (Cambridge Expedition to Suez Canal). Trans. Zool. Soc. XXII, pp. 231-286. 1936. Notes on some Decapod Crustacea of Bermuda. III. The larvae of the Palaemonidae. Proc. Zool. Soc. 1936, pp. 619-623. Kemp, S. W. 1922. Notes on Crustacea Decapoda in the Indian Museum. XV. Pontoniinae. Rec. Ind. Mus. XXIV, pp. 113-288. 1925. Notes on Crustacea Decapoda in the Indian Museum. XVII. On various Caridea. Rec. Ind. Mus. XXVII, pp. 249-343. Lebour, M. V. 1932. The larval stages of the Plymouth Caridea. IV. The Alpheidae. Proc. Zool. Soc. 1932, pp. 463-469. Ortmann, A. 1893. Decapoden und Schizopoden. Plankton Exp. Bd. II. Schmitt, W. L. 1935. Crustacea Macrura and Anomura of Porto Rico and the Virgin Islands. New York Acad. Sci. Scientific Surv. P. Rico, XV, pp. 125-227. Webb, G. E. 1921. The larvae of the Decapoda Macrura and Anomura of Plymouth. J. Mar. Biol. Ass. XII, pp. 385-417. Yokoya, Yu. 1931. On the metamorphosis of two Japanese fresh- water shrimps, Paratya compressa and Leander paucidens. J. College of Agric. Tokyo, XI, pp. 75-150. -= 3 MAR *938 ' 1 ! '1S4 ii- • : •iliV; "tJJW . ' ' ‘iv ■:>' -i l. i- '• .i . : Si - iiiiapii;; PMn# '* 1 .V 11 !« <3M ? :r<: BRITISH MUSEUM (NATURAL HISTORY) GREAT BARRIER REEF EXPEDITION. 1938-29 SCIENTIFIC REPORTS a9APR1940 VOLUME VI, No. 2 presented OX PHL YCTAENACHLAMYS LYSIOSQUILLINA GEN. AND SP. NOV,. A LAMELLIBRANCH COMMENSAL IN THE BURROWS OF LYSIOSQUILLA MACULATA BY MARY LEYBORNE POPHAM, B.Sc. Department of Zoology, University of Bristol • unur WITH TWENTY-ONE TEXT-FIGURES LONDON : PRINTED BY ORDER OP THE TRUSTEES OF THE BRITISH MUSEUM SOLD BY B. Quaritch, Ltd., 11 Uraftom Street, New Bond Street, London, W.l; Dui.au & Co., Ltd., 20 Dover Street, London, W. 1 ; Oxford University Prxb3, Warwick Sqvar s, London, E.C. 4 H.M. 3TATIONBRT o™, T.ONDOS. 3.W. 1 AND AT The British Mo9EOn (Natural History), Cromweli, Road, London, S.W.7 1939 [All rights reserved. Price Two Shillings and Sixpence. IllwlllttilP Jfc ■;? ® JiBl BUiflb' i '» — tm mm [Issued 2 5th March, 1939] Made and printed in Great Britain. 19 APR '940 PRESENTED OX PHLYCTAENACHLAMYS LYSIOSQUILLINA GEN. AND SP. NOT.. A LAMELLIBRANCH COMMENSAL IN THE BURROWS OF L YSIOSQ VILLA MACULATA BY MARY LEYBORNE POPHAM, B.Sc., Department of Zoology, University of Bristol. WITH TWENTY-ONE TEXT-FIGURES. CONTENTS. page 1. Introduction ............ 62 2. Habitat ............. 62 3. External Appearance and Habits ........ 62 4. Mantle and Shell ........... 65 5. Ctenidia 71 6. Labial Palps ............ 73 7. Alimentary Canal ........... 73 8. Foot ............. 77 9. Circulatory System .......... 77 10. Excretory System ........... 77 11. Nervous System 77 12. Genital System ........... 80 13. Description of Genus and Species ........ 80 14. Generic and Specific Characters ........ 82 15. Discussion 82 16. References ............ 83 vi. 2. 5 62 GREAT BARRIER REEF EXPEDITION 1. INTRODUCTION. A number of specimens of a commensal Lamellibranch with an internal shell and inhabiting the burrows of Lysiosquilla maculata were collected by Prof. C. M. Yonge during the course of the Great Barrier Reef Expedition. These differ in structure and habitat from any other Lamellibranch previously described, examination revealing that a new genus of the family Galeommatidae must be constituted to include this new species. The specimens collected were preserved, some in Bouin’s fluid, and others in 70% alcohol, and were handed over to the author for examination. One specimen was stained in borax carmine, and cleared in cedar-wood oil for examination of the entire animal. Serial transverse sections were prepared of another specimen and stained in Delafield’s haematoxylin and eosin. Two other specimens were dissected under a binocular microscope. Observations were made by Prof. C. M. Yonge on the habitat and appearance in life of this animal and his notes are incorporated in this paper, which has been prepared under his direction. 2. HABITAT. The tubes of Lysiosquilla maculata , one of the largest of the Stomatopoda, were common in the deep sand of the Sand Flat between the Sand Cay and the Thalamita Flat, as recorded by Stephenson, Stephenson, Tandy and Spender, in Vol. Ill, No. 2 of these Reports. The tubes were of great size with an internal diameter varying between 6 - 5 and 10 cm., and so extensive that an arm could be inserted up to the shoulder without the fingers coming in contact with the retreating occupant. The internal surface was very smooth. The external opening was much smaller, varying in diameter from 1 to 3 cm., but seldom being truly circular in outline. Fine sand from within the burrow was pushed over the edge to an extent of 4 to 5 cm., forming a ridged mound darker in colour than the surrounding surface sand. The overhanging ridge around the mouth of the burrows and the internal surface of these were very smooth, apparently owing to a viscid secretion produced by the occupant. Within the tubes, usually near to the opening, were frequently to be found one, or occasionally two, small animals with white, fleshy bodies. Closer examination revealed that these were lamellibranchs with greatly reduced, internal shells. In some cases the animals moved about, in others they were secured by one or two fine byssus threads projecting from the posterior tip of the elongated foot. 3. EXTERNAL APPEARANCE AND HABITS. The appearance of the living animal is shown in Text-fig. 1, and the side view of a preserved specimen in Text-fig. 2. The body length of the largest specimen was, in life, 2 -4 cm. The surface of the body is soft, white and partly translucent, and everywhere covered with small papillae which, when observed under a binocular microscope, have a slight brownish tinge. There are two conspicuous upward-directed tentacles (t) at the anterior end of the dorsal surface. These are deep yellow, almost orange in colour except at the base, where they are white. A number of smaller tentacles, similar in colour, are scattered irregularly over the surface of the body, and are more numerous in larger OX PHLYCTAEKACHLAMYS LYSIOSQUILLINA GEN. AXD SP. NOV.— POPHAM 63 specimens. They appear to represent enlarged papillae. All have considerable powers of expansion and contraction, especially the large anterior pair, which give the animals, at first sight, the appearance of nudibranchs. The shell valves (s) are completely enclosed within the mantle, and are greatly reduced. E Text-fig. 1. — View of the right side of the animal in life, x 2. e, exhalent opening ; fa, anterior portion of foot ; fp, posterior portion of foot ; I, inhalent opening ; s, shell ; t, tentacles. T Text-fig. 2. — Lateral view of the left side of a fixed specimen, x 12. p, papillae. Other lettering as in Text-fig. 1. The inhalent opening (i) is anterior, and consists of that part of the antero- ventral pallial opening that is dorsal and anterior to the base of the foot. A prolongation of the mantle forms a “ hood ” over the anterior part of the foot, the edges of the mantle being here slightly turned back. The exhalent opening (e) is situated on the end of a conspicuous rounded protuberance in the mid-dorsal line near the posterior end. The lips of this open and draw in rhythmically when the animal is in motion, although the opening never 64 GREAT BARRIER REEF EXPEDITION actually closes. The exhalent opening thus forms a more definite siphon than does the inhalent opening. The pedal opening, which is continuous with the inhalent opening, extends from this backwards about one-third of the distance along the mid-ventral surface, the combined openings being sufficient to allow projection of the foot and the ingress of the food stream anterior and dorsal to this. The foot can be completely withdrawn, and the aperture Text-fig. 3. — Ventral view of a fixed specimen. Text-fig. 4. — Dorsal view of a fixed specimen. X 12. c, ctenidium ; v, visceral mass. Other X 12. h, hinge. Other lettering as above, lettering as above. closed by the apposition of the mantle edges. Posterior to the pedal opening the mantle edges are fused as far as the exhalent opening. The foot is of considerable length when fully expanded, and consists of well-marked anterior (fa) and posterior (fp) regions. The anterior region is thicker, and is pushed out in a forward direction when the animal moves, resembling a long wedge about equal in length to the body. It possesses a creeping sole. The posterior portion is long and thin, with an opaque white area immediately short of the tip, which is sharply pointed. A byssus groove runs along the under surface and opens at the posterior end, from which byssus threads may project. Usually two of these were formed and planted between 30° and 90° apart when the animal was attached to the sides of the burrow. During movement the creeping sole advances and, at frequent intervals, the animal distends the mantle to the utmost extent and then, as a result of a convulsive movement which causes OX PHL YCTAEXACHLA M YS LYSIOSQUILLIXA GEX. AXD SP. NOV. — POPHAM 65 a backward projection of water through the exhalent opening, the body is drawn forward, thereby overtaking to some extent the foot, which continues to advance slowly and without interruption. The posterior end of the foot, like a fine thread, trails behind and is capable of very great extension, up to 3 cm. in a specimen 2’4 cm. long. In general external appearance the animal most nearly resembles Chlamydoconcha orcutti described by Bernard (1S97). Both have a globular form, with a white, semi- transparent mantle covered with papillae, are about the same size, have a foot with a creeping sole, and the position of the inhalent and exhalent openings is similar. But Chlamydoconcha occurs in stony clam beds, and the shell is different. 4. MANTLE AND SHELL. Histological examination of the mantle shows that beneath the epithelium it is composed of loose connective tissue, with very little structure (Text-figs. 12, m, and 20). It is thick, and the shell is completely embedded in it (Text-fig. 20). There are few blood- spaces, but a considerable amount of muscle. A definite muscular layer occurs below the internal epithelium bounding the mantle cavity (Text-figs. 12-19, ms). The arrangement of the muscle-fibres is irregular, and both circular and longitudinal muscles are present. This muscle layer must be responsible for the repeated contractions of the mantle observed during movements of the living animal. The arrangement is such that contraction of the muscle would reduce the volume of the mantle cavity and force water out through the exhalent opening. Distension is presumably due to ciliary action drawing in water when the exhalent siphon is temporarily closed. Fischer (1887) records similar convulsive movements, described as “ sphincter-like contractions of the mantle ”, in Chlamydoconcha orcutti which cause a current of water to pass over the gills. The muscle forms a thick band towards the posterior end of the mantle cavity (Text- fig. 18, ms). There is also a well-marked muscular region round the anterior tips of the shell- valves (Text-figs. 12 and 13). The muscle layer is absent from the mantle on the dorsal side of the visceral mass in the centre of the animal (Text-figs. 14-17). Strands of muscle also occur in the outer layer of the mantle, just external to the shell (Text-fig. 12, ms), but the chief muscular region is in the inner mantle layer between the shell and the mantle cavity (Text-fig. 20, ms). The mantle tissue between the shell and the mantle cavity is not so thick as that between the shell and the exterior. There is an irregularly-shaped cavity between the shell and the outer mantle layer (Text-figs. 13 and 20). The papillae which cover the surface (Text-fig. 2, p) also contain connective tissue, together with many dark-staining goblet-cells which open externally (Text-fig. 12a, mu). These are typical mucous glands and occur in the papillae all over the body. They are particularly numerous in the “ tentacles ” or larger papillae. The papillae do not appear to have a sensory function ; unlike the papillae in Chlamydoconcha (Bernard, 1897) they are not supplied by a nerve. Their function seems to be the secretion of mucus. The “ tentacles ” are similar in structure to the papillae, but they are more specialized. They have longitudinal muscle strands in the centre and are capable of considerable contraction and expansion. Histologically the cells have a more granular appearance, in the contracted condition of the fixed material, than the cells of the small papillae. 66 GREAT BARRIER REEF EXPEDITION They contain more mucous glands, irregularly arranged (Text-figs. 1 and 13, t). These small gland-cells also occur in Chlamydoconcha (Bernard, 1897), scattered over the surface of the mantle, but not on the papillae. The internal surface of the mantle lining the mantle cavity is generally smooth, but there is irregular folding in parts (Text-fig. 18). The shell measured 4 75 mm. x 3'0 mm. in a specimen 6'0 mm. in body length. It is fragile and has little protective function, as it only covers a small part of the dorsal Text-fig. 6. — a. Ligament, b. Right shell valve, c. Left shell valve, x 25. ct, cardinal tooth ; L, posterior lateral tooth. region of the visceral mass (Text-fig. 4, s). In shape it is equivalve, inequilateral, oval and flattened (Text-fig. 5). The valves are opposed only at the dorsal side in the region of the hinge. The ventral edges are always widely separated and can never be opposed, the visceral mass, ctenidia and foot, even in the contracted state, extending beyond them ventrally, anteriorly and posteriorly. In colour the shell is pure white over most of the surface, but becomes yellowish towards the umbo. The surface is smooth, with fine lines radiating from the umbo, and also lines, presumably indicating previous growth stages, following the curve of the circumference of the shell. The edge of the shell is smooth and entire. The umbo is situated one-third of the length from the posterior end ; it is circular and convex and occupies the summit of the valve (Text-fig. 5, u). The posterior OX PHL YC T A EX AC H LA M YS LYSIOSQUILLINA GEN. AXD SP. NOV.— POPHAM 67 end of the valve is short and rounded, while the anterior end is longer and more pointed, although the tip is blunt. The hinge-line is slightly curved. The hinge is situated in the region of the umbo below the summit of the valves. It is very poorly developed and the hinge teeth are small, blunt and rounded, with no means of interlocking. The hinge can, therefore, have little function. On the left valve the cardinal tooth is represented by a round knob in the centre of the hinge, and slightly anterior to the summit of the umbo (Text-fig. 6c, ct). On the right valve the cardinal tooth is probably represented by two knobs ; one immediately below the summit of the umbo is rounded and prominent, the other is anterior to the umbo and is broader and less prominent (Text-fig. 6b, ct). There are no anterior lateral teeth. The posterior lateral teeth are reduced to a ridge on the left valve, and to a socket between two ridges on the right valve (Text-fig. 6c, l). Text-fig. 7. — Lateral view of the left side of a fixed specimen with the mantle turned back, x 12. id, inner demibranch ; M, mantle ; od, outer demibranch ; other lettering as above. The shell valves are held together by an internal elastic ligament, inserted below the umbo. It is thick, dark brown in colour, and triangular in shape, with the base of the triangle situated posteriorly (Text-fig. 6a). The periostracum is well marked on the outer surface of the shell, and is also visible on the inner side at the edge of the shell (Text-fig. 20, po). The adductor muscles are poorly developed, and leave no impressions on the inner surface of the shell valves. Little movement of the shell valves occurs in life. The anterior adductor muscle is larger than the posterior, and is situated on the dorsal surface of the visceral mass over the oesophagus. It lies behind the labial palps and immediately above the cerebro-pleural ganglia (Text-fig. 9, aa). The posterior adductor muscle is small; it is situated at the posterior end of the visceral mass below the rectum and behind the heart (Text-fig. 9, pa). The morphological relationships of the shell and mantle call for comment. In the Lamellibranchia the free margin of the mantle normally consists of three folds (Text-fig. 68 GREAT BARRIER REEF EXPEDITION Text-fig. 8. — Ventral view with the mantle removed, x 12. al, ascending lamella. Other lettcrin as above. Text-fig. 9. — Dorsal view with the mantle removed, x 12. a, anus ; aa, anterior adductor ; k, kidney ; lp, labial palps ; mo, mouth ; pa, posterior adductor ; pr, pedal retractor ; R, rectum ; ve, ventricle. Other lettering as above. OX PEL YCTAEXACHLAM TS LYSIOSQUILLIXA GEX. AXD SP. XOV.— POPHAM 69 11a). The outer fold is always secretory in function; the middle fold is usually sensorv. and the inner fold forms the velum or pallial curtain where one is present (Yonge, 1936). In the animal here considered each valve is enclosed in the mantle, being completely surrounded by it. As described below, this condition has probably arisen owing to an overgrowth of the shell by the middle fold of the mantle. Other lettering as above. PO Text-fig. 11. — Relation of the mantle to the shell, a. Normal Lamellibranch. b. Hypothetical stage, c. Condition in Scintilla, D and e. Condition in PhlyctaenacJilamys lysiosquillina . ix, inner mantle lobe ; ml, middle mantle lobe ; ol, outer mantle lobe ; po, periostracum ; s, shell. The outer fold of the mantle remains secretory in function and produces the shell. This fold remains in the normal position immediately internal to the margins of the shell (Text-fig. 11, ol). The irregular cavity between the shell and this outer mantle fold is only present in some places, and is probably an artifact due to fixation. The periostracum, as in other Lamellibranchia, is formed at the base of the groove on its inner side, and extends over the edge of the shell as shown in Text -fig. 11, po. 70 GREAT BARRIER REEF EXPEDITION Differential growth of the middle fold of the mantle has probably taken place, first in a ventral direction and then dorsally over the outer surface of the shell (Text-fig. 11, ml). A hypothetical early stage of this overgrowth is shown in Text-fig. 11b. A further stage is found in Galeomma in which the mantle is reflected for a short distance over the anterior, posterior and ventral borders of the shell. In Scintilla , as shown in Text-fig. 11c, more of the shell is covered by the mantle (Deshayes, 1855). The final stage is shown in Text-figs. 11 d and e, and represents the condition found in this specimen. The over- growth is complete, and the middle mantle fold fuses in the mid-dorsal line with the middle fold of the opposite side. It also extends anteriorly and posteriorly. The fusion is complete posteriorly and dorsally except for the exhalent aperture. Anteriorly the Text-fig. 12. — Transverse section through the anterior end. x 33. f, foot ; mc, mantle cavity ; ms, muscle layer ; mu, mucous gland cell. Other lettering as above. Text-fig. 12a. — Papilla enlarged, x 200. co, connective tissue ; ep, epithelium. mantle folds form the sides of the inhalent siphon. The line of fusion cannot be detected in histological examination. The presence of a groove between the inner and middle mantle lobes on the inner side of the ventral extremity of the shell, and the irregular cavity between the shell and the mantle lying over it, support this interpretation. It also supplies an explanation for the presence of “ tentacles ”, as the middle fold normally bears sense-organs, although the sensory function appears to be lost in this case. The inner mantle fold forms the velum or pallial curtain in genera such as Pecten or Ostrea where the mantle edges are not fused ventrally. This is highly muscular, and performs extensive movements primarily concerned with controlling the entrance of water into the mantle cavity, e. g. in Ostrea, but modified in connection with swimming OX PHL TCTAENACHLAM YS LYSIOSQUILLIXA C4EX. AXD SI*. NOV.— POPHAM 71 in Pecten. Where fusion of the mantle lobes takes place, it is the inner mantle folds which fuse, and in the present case it is suggested that the inner fold grows inwards and fuses with the inner fold from the opposite side in the mid-ventral line, behind the pedal opening (Text-fig. 11. ix). There is again no histological evidence of this fusion, but the chief muscular region occurs in this part of the mantle, beneath the epithelium bounding the mantle cavity (Text-fig. 12, ms), and similar fusion of the mantle is found in verv manv Eulamellibranchia. 5. CTEXIDIA. The ctenidia each consist of two demibranchs formed of a descending or direct lamella and an ascending or reflected lamella. The ctenidia are inserted dorsallv behind the main part of the visceral mass. The axes are supported by the lesser reno-pericardial mass, the ascending lamellae of the outer demibranchs by the mantle, and the ascending lamellae of the inner demibranchs join in the mid-line below the reno-pericardial mass (Text-figs. 9 and 13). The inner demibranch extends farther forwards and downwards and has a more pointed anterior tip than the outer demibranch (Text-fig. 7, id). In fixed material the ctenidia appear to be plicated, the plicae occurring at regular intervals (Text-figs. 7-9). Histological examination shows, however, that there are no principal filaments, so the apparent plication may be due to fixation. X 33, Lettering as above. 72 GREAT BARRIER REEF EXPEDITION The structure of the ctenidia is very simple. The filaments are regular in arrangement with interfilamentary junctions formed by horizontal bars (Text-fig. 21, j). The lamellae of each demibranch are widely separated, with no interlamellar junctions. This unusual condition may be correlated with the distension and contraction of the mantle during life, and with the reduction of the shell and adductor muscles. In the fully distended condition the valves must be raised until they become almost horizontal in position, at the same time pulling out the ctenidia so that the lamellae of each demibranch are separated. This could not occur if they were connected by interlamellar junctions. When water has been Text-fig. 14. — Transverse section through the mouth, x 33. ap, anterior pallial nerve ; bg, byssal gland ; ip, inner labial palp ; op, outer labial palp ; pn, pedal nerve ; v, anterior end of the visceral mass. Other lettering as above. expelled through the exhalent siphon the volume of the mantle cavity is reduced, the valves lowered, and the ctenidia returned to their normal position with the two lamellae of each demibranch lying close together. According to Ridewood (1903), interlamellar junctions are present in all other members of the family Galeommatidae. 6. LABIAL PALPS. The two pairs of labial palps have the form of smooth flat blades with rounded tips (Text-fig. 9, lp), which extend laterally on either side of the mouth and curve down slightly OX PHL YCTA EX AC H LA M YS L YSIOSQUILLIXA GEN. AND SP. NOV. — POPHAM 73 into the mantle cavity on either side of the anterior tip of the inner demibranch (Text-fig. 14, ip, op). The free portion of the upper palp extends, therefore, to a position between the termination of the two demibranchs. The anterior median portions of the palps form the bps of the mouth, fusing in the mid-line and forming a solid oval band round the mouth. The opposed surfaces of the free portions of the palps are ridged, with a ciliated epithelium and many mucous glands. Cilia were not visible in fixed material on the outer surface of both pairs of palps, but they are probably present in life. 7. ALIMENTARY CANAL. The mouth is at the anterior end of the visceral mass, opposite the inhalent opening. It opens between the two lips formed by the median portion of the labial palps, being hidden by the upper pair (Text-fig. 14). The surface of the mouth is folded and the Text-fig. 15. — Transverse section through the cerebro-pleural ganglia. X 33. cc, cerebro- pleural commissure ; cg, cerebro-pleural ganglia ; cp, cerebro-pedal commissure ; d, digestive diverticula ; oe, oesophagus. Other lettering as above. 74 GREAT BARRIER REEF EXPEDITION ciliated epithelium contains many mucous cells. The lower lip is fused with the visceral mass. The mouth opens into the oesophagus, which is centrally situated below the anterior adductor muscle. It is oval in cross-section and similar in structure to the mouth, the epithelium being surrounded by a narrow band of circular muscle (Text-fig. 15, oe). It Text-fig. 16. — Transverse section through the stomach, x 33. cv, cerebro-visceral commis- sure; gs, gastric shield ; it, intestine ; li, ligament; ss, style-sac ; st, stomach ; te, testis. Other lettering as above. passes direct to the stomach, which is slightly more ventral in position and is surrounded by the digestive diverticula. This wide, oval cavity occupies a large part of the visceral mass ; it is lined with ciliated epithelium and has many mucous cells (Text-fig. 16, st). At the posterior end the stomach loses its oval shape and becomes irregular in outline. A large style-sac opens into the stomach to the right of the mid-ventral line. It extends ventrally into the mass of digestive diverticula (Text-fig. 16, ss), and is separate from the intestine. A smooth gastric shield covers the wall of the stomach opposite the opening of the style-sac (Text-fig. 16, gs). ON PHLYCTAENACHLAJIYS LYSIOSQUILLINA GEN. AND SP. NOV. — POPHAM 75 The digestive diverticula occupy most of the visceral mass, surrounding the alimentary canal from mouth to rectum (Text-fig. 15, d). In structure they consist of a mass of ramifying tubes, oval or round in cross-section, and clearly separated from each other. These tubules are lined with cubical cells containing granules, they open into ducts lined with flat ciliated epithelium, which unite and open through the ventral and lateral walls of the stomach, anterior to the style-sac. There are many separate openings, as noted by AL Text-fig. 17. — Transverse section through the rectum and the kidney, x 33. al, ascending lamella ; ca, axis of the ctenidium ; dl, descending lamella ; ic, interlamellar cavity ; pc, pericardium ; pp, posterior pallial nerve. Other lettering as above. Bernard (1897) in Clilamydoconcha , but they form three main groups, one ventral, and two lateral. The intestine leaves the stomach just anterior to the opening of the style-sac, and to the right of the mid- ventral line. It coils about through the mass of digestive diverticula, remaining ventral to the stomach throughout its course (Text-fig. 16, it), except during the last coil, when it passes forwards and upwards to the right of the stomach. Behind the stomach the intestine passes upwards in a loop through the posterior end of the gonad, on a level with the middle of the stomach. It gradually becomes more dorsal, finally merging into the rectum. The intestine has no typhlosole. 76 GREAT BARRIER REEF EXPEDITION R OX PEL YCTAEXACELA3I YS LYSIOSQUILLIXA GEN. AND SP. NOV.— POPHAM 77 The rectum is a straight circular tube. It has no mucous glands ; this is possibly correlated with the powerful expulsion of water through the exhalent siphon rendering the elaboration of firm faecal pellets unnecessary. The rectum passes through the ventricle of the heart, over the posterior adductor (Text-figs. 17 and 18. r), and opens by a dorsally directed anus into the mantle cavity, below the exhalent siphon (Text-fig. 19, a). The whole of the alimentary canal, with the exception of the region bearing the gastric shield, is strongly ciliated. 8. FOOT. Histologically the foot is a compact mass of muscle with many blood spaces in the centre. It is also well supplied with nerves (Text -figs. 12-16, fa, fp). The ventral surface of the foot, together with the sides of the byssal groove, is ciliated. These cilia are probably responsible for the slow gliding motion of the foot in life. In the anterior portion of the foot there are typical mucous glands, the mucus produced lubricating the creeping sole of the foot. They are most numerous ventrally, but they also extend up the sides of the foot at the anterior tip ; farther back they are confined to the ventral side, and eventually mingle with the byssal gland-cells. The byssal gland consists of dark-staining cells concentrated round the inner portion of the byssal groove, but extending for a considerable depth into the pedal mass. The gland is dense in the central portion of the foot, and extends above the byssal groove throughout the posterior portion of the foot, except for the extreme posterior tip. The paired pedal retractor muscles are inserted into the. posterior portion of the shell, just anterior to the anus, and immediately above the posterior adductor muscle (Text-figs. 9 and 18, pr). They are most widely separated at the posterior end. They pass forwards on either side of the heart and kidneys (Text-fig. 17, pr), to either side of, and above the visceral ganglia. The two muscles join below the intestine, and pass into the substance of the visceral mass at its posterior end. They pass down either side of the posterior part of the gonad, and into the muscular mass of the foot. The pair of pedal protractor muscles are slender. They pass forwards on either side of the oesophagus, remaining below the anterior adductor muscle and the cerebro-pleural commissure, and enter the anterior portion of the foot in front of the cerebro-pleural ganglia. 9. CIRCULATORY SYSTEM. The pericardial cavity is posterior and dorsal to the visceral mass. It is situated between the two pedal retractor muscles, in the mid-line immediately behind the kidneys (Text-fig. 17, pc). It contains the heart, which consists of a ventricle and two lateral, thin-walled auricles (Text-fig. 10, au). The blood-vessels are not well defined, except near the heart, where anterior and posterior vessels leave the heart in the mid-line. Two lateral vessels enter the auricles close to the mid-line, they come from the visceral mass. 10. EXCRETORY SYSTEM. The kidneys are small. They are situated just anterior to the heart and, with the pericardium, form a distinct mass separate from the main portion of the visceral mass, and posterior and dorsal to it (Text-fig. 10, k). At the anterior end there are two distinct vi. 2. 6 78 GREAT BARRIER REEE EXPEDITION organs, one on either side of the rectum and separated by the anterior end of the pericar- dium. As the pericardium narrows posteriorly, the two kidneys enlarge towards the mid-line. They end posteriorly in a point beneath the posterior adductor muscle. Anteriorly they extend as far as the posterior end of the gonad. A ciliated duct from each kidney opens into the pericardium immediately above the pedal retractor muscle, on either side of the rectum, and below the posterior end of the gonad. The opening from the kidneys into the mantle cavity has not been observed. The pericardial gland is situated on the lateral walls of the pericardium, and also covers the lateral walls of the anterior end of the ventricle. It extends, at the level of the axes of the ctenidia, throughout the greater part of the pericardium, and ends above the posterior face of the visceral ganglia, and opposite the auricles. It consists of a dense mass of cells which contain granules and have large nuclei. 11. NERVOUS SYSTEM. The nervous system is normal and the ganglia are large. The cerebro-pleural ganglia are situated behind the mouth, on either side of and slightly above the oesophagus (Text-fig. 15, cg), they are joined by a large cerebral commissure, which passes over the oesophagus and under the anterior adductor muscle (Text-fig. 15, cc). Large pallial nerves leave the anterior face of the cerebro-pleural ganglia, and pass forwards on either side of the mouth, becoming slightly more dorsal in position and farther apart (Text-fig. 14, ap). They lie in the mantle, close under the shell, and become farther apart as the anterior tips of the shell valves separate. Towards the tip of the shell the pallial nerves divide up into many small nerves, which spread out in the tissue of the mantle inside the shell. The main branch of the nerve is situated centrally beneath the shell. At the extreme anterior end the small nerves lie just below the muscle layer that borders the mantle cavity. The cerebro-pedal commissures leave the posterior ventral face of the cerebro-pleural ganglia, and pass vertically downwards into the foot between the mass of digestive diverticula and the wall of muscular tissue that bounds them (Text-fig. 15, cp). Upon reaching the muscular mass of the foot the nerves pass backwards in the centre of the foot on either side of the median line and into the pedal ganglia at their anterior face. The pedal ganglia are situated in the central portion of the foot at the approximate dividing line between the anterior and posterior portions. They lie between the byssal gland at the ventral side of the foot and the gonad. The ganglia are large and joined together in the mid-line. Large pedal nerves pass forwards from the ganglia on either side of the mid-line into the anterior portion of the foot, and divide up into many smaller nerves. The whole of the anterior portion of the foot' is richly ennervated (Text-fig. 15, pn). Posterior pedal nerves leave the posterior face of the pedal ganglia. They are smaller than the anterior pedal nerves and supply the posterior portion of the foot. Like the anterior pedal nerves, they divide up into many small nerves. The cerebro-visceral commissures leave the posterior face of the cerebro-pleural ganglia ; they pass back on either side of the oesophagus, becoming farther apart and more ventral in position. They are smaller than the cerebro-pedal commissures or the anterior pallial nerves. They pass between the mass of digestive diverticula, and the muscular OX PHL YCTA EXACHLA M YS LYSIOSQUILLIXA GEX. AND SP. NOV.— POPHAM 79 wall of the visceral mass, and are at the same level as the axes of the ctenidia (Text-fig. 16, cv). They he close to the wall of the stomach at its posterior end, approach the median fine behind it and then extend beneath the dorsal portion of the gonad. They continue on either side of the heart, above and outside the pedal retractor muscles, and pass towards the mid-line beneath the posterior end of the heart. They enter the visceral ganglia at their anterior face beneath the rectum. Text-fig. 20. — Transverse section through the mantle and the shell, x 75. im, inner layer of the mantle ; n, nacreous layer of the shell ; om, outer layer of the mantle ; pm, prismatic layer of the shell. Other lettering as above. The visceral ganglia are the largest of the ganglia and are situated behind the main visceral mass, beneath the pericardium and excretory organs. They lie close together, and are joined in the mid-line into one large ganglionic structure. A large branchial nerve passes from the external lateral side of each visceral ganglion into the main axis of one of the ctenidia, continuing along this and sending branches into the descending lamellae of each demibranch. A nerve passes from the ventral face of each ganglion into the ascending lamellae and along the axes, sending branches into each lamella. A nerve also leaves the posterior face of each ganglion. These two nerves separate and pass beneath the pedal retractor muscles and the pericardium. Posteriorly 80 GREAT BARRIER REEF EXPEDITION they become farther apart and pass into the tissue of the mantle inside the shell, where they divide up into many small nerves. They remain below the posterior adductor muscle. There are three main branches which lie close under it, one in the mid-line, and two laterally (Text-fig. 18, pp). The whole of the mantle inside the shell is richly innervated (Text-fig. 20, pp). 12. GENITAL SYSTEM. The specimen sectioned was a male, and the testis, composed of ramifying tubules, contained sperms in all stages of development. The testis extends forwards above the foot, to the level of the pedal ganglia. At the anterior end it is confined to the ventral side of the visceral mass, and is spread out over the ventral surface of the digestive diverticula (Text-fig. 16, te). The mass of the gonad is small compared with that of the digestive diverticula. Behind the style-sac the testis spreads out dorsally and composes most of the posterior part of the visceral mass. It surrounds the hinder, part of the intestine. At the posterior end of the visceral mass the testis becomes more dorsal in position, and forms a compact mass surrounding the rectum. The testis opens by a short ciliated funnel into the mantle cavity on either side. The opening leads into the supra-branchial cavity in the region above the inner demibranch, immediately over the cerebro- visceral commissure, and just anterior to the opening of the excretory organ into the pericardium. 13. DESCRIPTION OF GENUS AND SPECIES. According to Pelseneer’s classification (Pelseneer, 1911), the Order Lucinacea is divided into two groups according to whether the ctenidia have one or two demibranchs : Lucinacea. Double gill lamellae Leptonidae Erycinidae Galeommatidae Pythina Galeomma Scintilla Ephippodonta Chlamydoconcha Kellia Lasaea Internal shell Single gill lamella I Montacutidae I ' I Montacuta Jousseaumiella ( Sciorberetia [Entovalva. The Erycinidae are distinguished by the presence of three pallial openings, whereas the members of the other three families have two. The Erycinidae and Montacutidae are hermaphrodite, with incubation of the embryos in the mantle cavity. In the Leptonidae and Galeommatidae the sexes are separate, and the mantle has papillae. The distinction between the two families lies in the position of the shell in relation to the mantle. In the ON PHL YGTAENACHLAM YS LTSIOSQUILLINA GEN. AND SP. NOV. — POPHAM 81 Leptonidae the shell is always external, blit in the Galeommatidae the mantle covers the shell to a varying degree. In Pythina (Pelseneer, 1911) the mantle is only reflected over the extreme edge of the shell. In Galeomma and Scintilla (Fischer, 1887) the mantle is reflected for some distance over the edges of the shell valves, while in Ephippodonta (Tate, 1887) and Chlamydoconcha (Bernard, 1897) this reflection of the mantle has advanced to such an extent that the shell is completely internal. In the Erycinidae the shell is always external, but a similar development to that in the Galeommatidae has taken place in the Montacutidae. Montacuta has an external shell with no reflection of the mantle lobes. J ousseaumiella (Bourne, 1906) also has an external shell, but in Sciorberetia (Bernard, 1895) and Entovalva (Ohshima, 1930) the shell is completely enclosed in the mantle as it is in Ephippodonta and Chlamydoconcha. Modification in the position of the shell valves has also occurred in this order, and may be related to the covering of the shell by the mantle. In Kellia and Lasaea, in which the shell is external, the foot can be withdrawn and the shell valves closed. In Scintilla the valves gape slightly (Deshayes, 1855), but less than in Galeomma, in which there is a permanent ventral gape, and in these species there is not only a gape, but the foot and part of the visceral mass extend permanently beyond the shell. In Ephippodonta the valves are semicircular, wide apart, and cannot be depressed to an angle of less than 70° The animal is flat (Tate, 1887). The extreme case is found in Libratula, one of the Galeom- matidae, in which the valves are normally perfectly horizontal and cannot be depressed to an angle of less than about 90° (Pease, 1855), but in this case the shell is external. All members of the Lucinacea have an anterior inhalent opening and a posterior exhalent opening, and throughout the group the foot has a creeping sole. In the species here examined for the first time there are two gill lamellae, two pallial openings, and the shell is completely enclosed in the mantle, which has papillae. There is a permanent gape between the shell valves, and the foot and part of the visceral mass extend permanently beyond the shell. The sexes are separate. It must therefore be included in the Galeommatidae and grouped with Ephippodonta and Chlamydoconcha, to which it is most closely related. In external appearance it most closely resembles Chlamydoconcha, but the shell is widely different (that of Chlamydoconcha being sword- shaped), the buccal hood is less prominent, the “ anterior orifice ” is absent, the papillae on the mantle are more numerous, the foot is longer and thinner, the adductor muscles are present, although reduced, and the habitat is different, the new species occurring in the burrows of Lysiosquilla, while Chlamydoconcha is found in stony clam-beds fixed to rocks or pebbles. The animal differs from Ephippodonta in the form of the shell, that of Ephippodonta being semi-circular and covered in spines, with two bifid cardinal teeth in each valve. (These do not interlock.) The mantle lobes of Ephippodonta are free all round except dorsally, and the anterior margins of the mantle lobes are expanded and form a funnel. The foot is disc-shaped and very large, with a broad locomotory surface. There is also a difference in habitat. Ephippodonta, which was collected at Edithburg, Yorke Peninsula, South Australia (Tate, 1887), lives in the mud-formed burrows of a shrimp, sheltering beneath large stones between tide-marks. The shell of the new species more closely resembles that of other members of the Galeommatidae, and is also similar to the shell of Sciorberetia (Bernard, 1895) and Entovalva (Anthony, 1916 ; Ohshima, 1930), both members of the Montacutidae, 82 GREAT BARRIER REEF EXPEDITION It will be seen from the above that although this species must be included in the Galeommatidae, and is related to Ephippodonta and Chlamydoconcha, it differs to such an extent from both of these that it is reasonable to establish a new genus for its inclusion. 14. GENERIC AND SPECIFIC CHARACTERS. Phlyctaenachlamys lysiosquillina gen. and sp. n. Habitat.- — Occurs in the burrows of Lysiosquilla maculata. Length of largest specimen in life : 2-4 cm. Particulars of Type Specimen in the British Museum (N. H.) : Length 6-0 mm. Shell, 475 x 30 mm. Mantle covered with papillae; these are not sensory, but have many mucous glands. Two large retractile tentacles at the anterior end ; these are enlarged papillae. Mantle forms a hood over the anterior portion of the foot. Well-developed muscle layer in the mantle. Exhalent opening situated on a round protuberance posteriorly. Shell completely enclosed in the mantle ; reduced, smooth, white and fragile, oval in shape and inequilateral. Gape between the shell valves ventrally ; the mantle, foot, visceral mass and ctenidia extend permanently beyond them. Hinge teeth reduced. Internal elastic ligament. Adductor muscles reduced. Foot in two parts ; anterior portion wedge-shaped ; posterior portion long and thin. Foot contains many mucous glands, as well as a byssal gland. Ctenidia have no interlamellar junctions. Style-sac separate from the intestine. Intestine has no typhlosole. Rectum has no mucous glands. Sexes separate. 15. DISCUSSION. Phlyctaenachlamys lysiosquillina is a commensal from the burrows of Lysiosquilla maculata. The reduction of the shell, the covering of the shell by the mantle and the associated reduction of the hinge-teeth and the adductor muscles are adaptations in relation to this commensal mode of life. In the Lucinacea an interesting series can be traced in the development of commen- salism culminating in parasitism. Kellia, Galeomma and Chlamydoconcha are examples of the free-living members ; Ephippodonta is commensal in the burrows of a prawn (Tate, 1887 ; Woodward, 1893) ; Lepton squamosum is commensal with Upogebia stellata ; Lepton clarkia and Mysella bidentata with Phascolosoma elongatum (Salisbury, 1932) ; Mysella bidentata also with Ophiocnida brachiata (Winckworth, 1924) and Montacuta ferruginosa with Echinocardium cor datum. None of these is attached to the animal with which it is associated, and this is also the case in Phlyctaenachlamys lysiosquillina. Montacuta sub- striata, however, lives attached by byssus threads to the spines of Spatangus purpureus (Salisbury, 1932) ; Entovalva semperi lives attached to the body of a synaptid Protankyra bidentata (Ohshima, 1930) ; Joussecmmia in the basal chambers of the corals Heteropsammia and Heterocyathus (Bourne, 1906), while the extreme case is found in Entovalva mirabilis, which is an internal parasite in the gut of a holothurian (Voeltzkow, 1891). Phlyctaenachlamys lysiosquillina produces byssus threads by which it may be attached to the sides of the burrow. The foot has a creeping sole, and movement is assisted by the unusual method of forcing water out from the mantle cavity by sudden violent con- tractions of the mantle. There are various modifications in relation to this form of movement, namely the development of an extensive layer of muscle in the mantle, and the 83 OX PUL YCTAEXACHLAM YS LYSIOSQVILLINA GEN. AXD SP. NOV. — POPHAM loss of the interlamellar junctions from the ctenidia. These interlamellar junctions are present in both Chlamydoconcha (Bernard. 1897) and Ephippodonta (Ridewood, 1903) and throughout the group, although Bernard records similar, although presumably less extensive, movements in Chlamydoconcha. This form of ctenidium will offer less resistance to the stream of water than one with a greater rigidity, and will be better able to withstand the considerable strain put upon it, due to the increase of pressure in the infra branchial cavity when the mantle contracts. The absence of mucous glands in the rectum is probably also associated with this movement of water through the mantle cavity. Particles of faecal matter will be shot out through the exhalent opening. There is thus no necessity for the formation of firm faecal pellets, because, owing to the violent expulsion of water, there will be less danger of the mantle cavity silting up. Text-fig. 21. — Ctenidium. x 150. ci, cilia; fi, filament; ic, interlamellar cavity; j, inter- filamentary junction. The presence of a direct food current in an anterior to posterior direction, which represents the movement of the inhalent aperture from its primitive position, ventral to the exhalent aperture at the posterior end, to the anterior end, occurs only in the Order Lucinacea. It is clearly correlated with the absence of the burrowing habit in these animals, and probably with their mode of progression by means of a creeping sole. It is certainly mechanically efficient, as the ctenidia will act as “ strainers ” in the line of a direct current of water passing through the mantle cavity. 16. REFERENCES. Anthony, R. 1916. Contribution a l’etude de V Entovalva (Synapticola) perrieri, Malard. Mollusque acephale commensal des Synaptes. Arch. Zool. Exper. Gen. LV, pp. 375-391 . Bernard, F. 1895. Sciorberetia australis type nouveau de Lamellibranche. Bull. Sci. France Belg. XXVII, pp. 364-395. 1897. Chlamydoconcha orcutti. Ann. Sci. Nat., Zool. (8) III, pp. 220-250. Bourne, G. C. 1906. On Jousseaumia. Cey. Pearl Oyster Fish. Rep., Roy. Soc. London, V, pp. 243-266. Dall, W. H. 1884. Chlamydoconcha. Science, IV, pp. 50-51. Deshayes, G. P. 1855. Sur le genre Scintilla. Proc. Zool. Soc. London, 1855, p. 171. Fischer, P. 1887. Sur un nouveau type de Mollusque. Chlamydoconcha. Man. de Conchyl. et de Palaeont. Conchyl., 1887, pp. 201-206. 84 GREAT BARRIER REEF EXPEDITION Ohshima, H. 1930. Preliminary Note on Entovalva semperi sp. nov., a Commensal Bivalve Living Attached to the Body of a Synaptid. Annot. Zool. Jap. XIII, pp. 25-26. Pease, W. 1865. Descriptions of New Genera and Species of Marine Shells from the Islands of the Central Pacific. Proc. Zool. Soc. London, 1865, pp. 512-517. Pelseneer, P. 1911. Les Lamellibranches de 1’Expedition du Siboga. Partie Anatomique. Siboga- Expeditie. Mon. Lilia. Ridewood, W. G. 1903. On the Structure of the Gills of the Lamellibranchia. Phil. Trans. Roy. Soc., b, CXCY, pp. 147-284. Salisbury, A. E. 1932. On Lepton squamosum and Montacuta ferruginosa and some other molluscs observed at the Salcombe Estuary, Devon. Proc. Malac. Soc. London, XX, pp. 100-103. Tate, R. 1887. Descriptions of some New Species of Marine Mollusca from South Australia and Victoria. Trans. Proc. Rep. Roy. Soc. S. Aust. XI, pp. 62-75. Voeltzkow, A. 1891. Entovalva mirabilis eine schmarotzende Muschel aus den Darm einer Holothurie. Zool. Jahrb. Abt. Syst. V, pp. 619-628. Winckworth, R. 1924. Specimens of Molluscs and Other Animals associated with them Found in a Day’s Digging at Salcombe. Proc. Malac. Soc. London, XVI, p. 1. Woodward, M. F. 1893. Ephippodonta MacDougalli. Proc. Malac. Soc. London, I, pp. 20-25. Yonge, C. M. (1936) The Evolution of the Swimming Habit in the Lamellibranchia. Mem. Mus. Roy. d’Hist. Nat. Belg. (2), Fasc. 3, pp. 77-100. 19 APR 1940 PRESENTED BRITISH MUSEUM (NATURAL HISTORY) GREAT BARRIER EEEF EXPEDITION 5.3°° 1928-29 SCIENTIFIC REPORTS VOLUME VI, No. 3 MADREPORARIA, H YI) ROCO R ALLIN A E , HELIOPORA AND TUBEPORA BY CYRIL CROSSLAND. D.Sc. (1878-1943) WITH ONE TEXT-FIGURE and FIFTY-SIX PLATES LONDON : PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM SOLD BY B. Quaritch, Ltd., 11 Grafton Strbet, Nsw Bond Strut, London, TV. I H.M. Stationery Office, London, S.W. 3 AND AT Ths British Museum (Natural History), Cromwicll Eoad, London, S.W. 7 1952 [All rights reserved _ Price Three Pounds, Ten Shillings [Issued 22nd October, 1952 j MADREPORARIA, H YDROCOR ALLIN AE, HELIOPORA AND TUBIPORA BY CYRIL CROSSLAND. D.Sc. (1878-1913). WITH ONE TEXT-FIGURE AND FIFTY-SIX PLATES. EDITOR’S NOTE Cyril Crossland undertool- to report upon the coral collections of the Great Barrier Reef Expedition and the material was sent to him just before the outbreak of war in 1939. The task, interrupted by illness, took him three years, but he had completed it before his death early in 1913. Thanks to the kindness of the staff of the Zoological Museum of the University of Copenhagen, and especially of Dr. P. L. Kramp, his typescript and notes, and the collections were carefully preserved until communications were re-opened at the end of the tear. Although the report was complete in most respects, no illustrations had been prepared. The difficult task of choosing material to illustrate to the best advantage the points made by the author and of supervising the prepara- tion of the plates has been carried out with meticulous care by Capt. A. K. Totton, M.C. and Mr. E. White of the Zoological Department, British Museum (Natural History) ; they have also been responsible for the final checking of the text and seeing it through the press. It has been their aim to preserve, as far as possible, the integrity of the original text and any necessary departures from it, except the most trivial, have been indicated by notes bearing Capt. Tottoris initials. In a covering letter, found with the typescript, the author stressed the importance of good and numerous illustrations for any paper dealing with corals. He also urged that his report should be accompanied by actual photographic prints, or collotype reproductions, in order that details not mentioned in, or not clear from, the text could be made out by examination under a lens. It is hoped that the number of illustrations now presented with his report would not have disappointed him. It is also hoped that in the future no one will again be com- pelled to rely, as Crossland was, on illustrations alone, but that there will be freedom to consult the actual specimens or photographs prepared specifically for any particular investigation. CONTENTS PAGE 1. Foreword: Cyril Crossland. By Dr. A. F. Bruun .... 86 2. Introduction ............ 88 3. List of Species 89 4. Description of Species 101 5. References . . . 251 6. Index of Genera and Species 255 vi, 3. 7 8G GREAT BARRIER REEF EXPEDITION FOREWORD CYRIL CROSSLAND, STUDENT OF CORALS “ The foundation of my love of science is my love of beauty.” This is the first sentence of some autobiographical notes left by Cyril Crossland, and to which he has given the heading “ How I became a Coral Reef Man.” Crossland died in his 65th year on 7th January, 1943, in Denmark. A general outline of his life has been given elsewhere (Vidensk. Medd. fra Dansk Naturhist. Forening, CVI, Copenhagen 1943). Here I shall try to give a picture of Crossland as a student of corals and coral reefs. It is based on the autobiographical notes mentioned and letters from the time when I first met him in Tahiti (or Tahaiti, as he always maintained was the correct spelling) in 1928, and not least from memories of a close friendship developed during his last years in Denmark from 1938 until his untimely death. Crossland had a very high opinion about what ought to be the standard of scientific papers and was most self-critical. This may to some extent account for the fairly restricted number of Coral papers he produced, while he himself, modestly admitting “ it has been very little,” accounts for it by the force of circumstances : ft for, except when in Tahaiti, when I was greatly hindered by ill health, corals and coral-reefs have been secondary to the work for which I drew the pay, and I have conscientiously kept them secondary.” The number of coral papers published in his later years, of which the present was finished just before the final illness that led to his death, indicates that this was the real reason. No better example of his patient study of detail can be given than his account of the Forskal Collection of Corals, so important for nomenclature. By a wonderful coincidence he had all the qualifications required for taking up that study ; he had a wide knowledge of Latin and Arabic, he knew all the places where Forskal collected his specimens, and had some twenty years of personal knowledge of the coral reefs of the Red Sea. Without being a specialist on corals I venture to say that that paper will range among the classics of coral taxonomy. But systematics was but one of the tools for Crossland in his study of the general coral reef problems which right from the start in 1900 thrilled his imagination when he first saw living corals at Prison Island, in Zanzibar harbour. “ The great reef of the east coast from which the roaring surf has been heard at Dunga, in the centre of the island, 10 miles inland, about 15 from the actual reef edge, was my Mecca. I would see the great waves from Chuaka rise up white, apparently standing still like the sail of a boat, before falling in foam. Even after I had settled at Chuaka it was some time before I could get out to the reefs, in calm weather at low tide, and walk on the edge. To my great surprise I found it all quite dead.” From his studies of the reefs of Zanzibar, Equatorial East Africa, the Red Sea, Tahiti and some other places Crossland arrived at the astonishing conclusion, suggesting that “ the age of corals is past, not only in Tahiti and elsewhere, and the cliff which generally surrounds atolls, at about 50 fathoms below the surface ” (J. Linn. Soc. Lond. XXXVI, 1928). He was fully aware that the reefs he had seen appeared to be unlike those described MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAXD 87 by others, but be was convinced that they had to be taken into account. At the same time he always stressed the point that no coral reef area and almost no coral reef has yet been completely examined ; this was not meant to underrate good work done by other scientists, but from his own experience he knew what an enormous amount of work is still left. As an example he mentioned the East African reefs, which he examined “ when I was young (very young ) in 1900, and, in the succeeding 39 years no one else has been there to verify, disprove, or to complete my description of those vast dead reefs, nor are they mentioned in any recent literature, though they are a really stupendous phenomenon of obvious importance to the history of coral growth.” Among the principal omissions from many accounts of reefs he found the most obvious to be the description of the outer slope, beneath and outside the breakers, as far as the bottom is visible ; he was aware of the difficulty and danger of the job, though he thought the danger more apparent than real. He found the observations from the Funafuti expedition most valuable, but to have any general value they must be multiplied many times. He would also claim that for most areas we have only incidental references to what are the dominant genera of corals, mentioning vast and wholly unexplained differences such as the dominance of Pontes ( Synaraea ) and leafy Pavona in the Maldives, where the leafy species, usually so much more abundant, are absent, and the rarity of all colonial Fungiidae in the northern Red Sea, and several others. “ Nature has so many tricks to play upon us, only some of which have been discovered that, as in all other investigations, no one man’s work can be anything but a beginning.” He considered himself the last of the pioneers, who work alone with no elaborate equipment, “ the present fashion being to go out in parties, which may degenerate into crowds, a distraction and hindrance to the leader.” “ Simplicity, Scepticism, and Patience ; Walking, Cycling, and Canoe Paddling, have been, for the most part, my own methods, and they have served me well.” At the same time he fully admitted that a coral reef man cannot work in isolation , the problems are too complex. The biologist, geologist, oceanographer, and meteorologist, etc., should work together and hold discussions on the spot. “ Team work is thus even more necessary in coral reef work than in others, but it is not necessary that the team should work simultaneously in a crowd, hindering each others’ movements, and leaving little privacy for that quiet thought, apparent idleness, which is so essential for anything but routine work.” For the future he wanted much more accumulation of “ mere ” descriptions and “ mere ” observations, with a programme of experiment running parallel, general theory to grow step by step as progress was made ; he thought that theory has outrun observation to a large extent. It is my hope to have given some glimpses of Crossland as the Coral Reef Man behind his printed works. To have been with him, he wearing a native loin cloth, a pareo, and a straw hat, in a leaking outrigger canoe on the lagoons of Tahiti, to have listened to his enthusiastic explanations of the wonderful sights in his water telescope I reckon as a very great privilege ; a still greater to have known this noble and gentle character. ANTON F. BRUUN. 88 GREAT BARRIER REEF EXPEDITION INTRODUCTION. This is the largest collection of corals brought from any one area, containing as it does 174 species, divided among 54 genera. Of these, 30 species are new, 2 of Bernard’s have been given names, and there are several curious varieties, 3 of which have been given new names. This is evidence, not only of the richness of the fauna but of very careful collecting, since several species, which have not been seen for many years, are included. Yet it is clear that the collection does not include all the species which inhabit the Great Barrier Reef area ; even assuming that Brook and others have made many synonyms, there are species described by them which have not been found again, and some of those described by T. W. Vaughan from Murray Island do not occur in this collection. This latter difference probably indicates a real difference in the faunas, the Murray Islands being about 10° S. and Low Isles 16° 20' S., but the difference in their positions with respect to the Barrier has probably an even greater influence. Dr. and Mrs. Stephenson and others remark of Yonge Reef (Vol. Ill, No. 2, p. 83) : “ The coral and alcyonarian species are unlike those with which we had become familiar, and although closer study showed that many of them represent modified growth-forms of Low Isles species, or forms which in the latter place are limited to the windward side, others were actually new, and everything looked a little different.” Low Isles is distinctly muddy. The quite unexplained difference in the faunas of different reefs is frequently referred to in what follows. My experience corrobor- ates Umbgrove’s, who (1939, p. 12) writes : “ We are only in the very beginning of having an idea of the true distribution of species which abound in one reef but seem absent in some other reefs.” I continue this quotation under Acrohelia Tiorrescens, Montipora foliosa and M. ramosa, but the fact applies to other species also. I have noted of some species that they are absent from the Red Sea or Tahaiti, but proving a negative is, of course, always a doubtful business. For instance in my three years in Tahaiti I found only one specimen each of Astreopora, Pachyseris, and Echinophyllia, the latter the extraordinary specimen illustrated in Plate II (1935). In the Red Sea I specially looked for the Cymodocea with strong stems rising from rhizomes and bearing tufts of broad leaves, which is such an ubiquitous feature of the reefs of Zanzibar and neighbour- hood, but it was some years before I found one small patch on the Sudan coast. Years later I found it abundant at Ghardaqa, in the north. Here is a list of Murray Island species described by T. W. Vaughan, but not found by the Great Barrier Reef Expedition : — Seriatoporidae : Seriatopora angulata, possibly the same as S. hystrix : Acroporidae : Montipora turgescens (the only species of Montipora not found) ; Acropora decipiens ; A. pectinata ; A. spicifera ; A. sarmentosa ; A. plicata ; A. murrayensis V. (possibly the same as A. rosaria) ; A. squarrosa ; A. syringodes. MADREPOBARIA, HYDROCOR A LLTNAE, HELIOPORA AND TUBIPORA— CROSSLAND 89 Thamnastreidae : Psammocora gonagra. Fungiidae : Fungia sp. aff. condnna. Poritidae : Porites murrayensis Vaughan ; P. may&ri Vaughan ; P. viridis Gardiner. Faviidae : Coeloria stricta, synonymous with C. daedalea ; Leptoria gracilis , probably the same as L. phrygia ; Hydnophora exesa, possibly the same as H. microeonos. Here is the list of species, with previous records, if any, from the Barrier Reef region, the recorder’s name being indicated by initials, as in the list of abbreviations below. Note Cynarina savignyi, which has been, so far, found only in the Gulf of Suez, and turns up again now on the other side of the world. On the other hand, the fewness of the species common to the Great Barrier Reef and the Red Sea, among the Fungiidae and Perforata, is somewhat surprising. There is, for a considerable part of the fauna, a Pacific region, beginning in Malaysia and dying out beyond Fiji (the continental limit), considerably reduced in Samoa, still further in Tahaiti,* with relics only in the Marquesas. LIST OF SPECIES COLLECTED BY MEMBERS OF THE GREAT BARRIER REEF EXPEDITION, WITH A NOTE OF PREVIOUS RECORDS OF THEM FROM THE GREAT BARRIER REEF AND OF THEIR WIDER DISTRIBUTION. Other recorders from Gt. B. Reef.f Wider distribution. Page. Madreporaria [Caryophylliidae] Caryophylliinae Trachyphyllia geoffroyi M. Northern Red Sea, Indo-Pacific . 101 Heterocyathus aeq uicostat us to Philippines ; Indian . 102 [Thecocyathinae] Thecocyathus minor . Ocean, including Natal and Equatorial East Africa ; Persian Gulf ; W. Australia ; Pacific to Philippines. Loyalty Is. . 103 [Eusmiliinae] EuphylUa glabrescens m. Indian Ocean, including Gulf . 104 E.fimbriula M. of Aden ; Pacific to Samoa. Red Sea, Indo-Pacific (western). . 105 Flabellidae Flabellum rubrum Indian Ocean, including Cape ; . 105 F. vacuum sp. n. New Zealand ; Pacific as far as Palau Islands. . 106 * I adopt this spelling after Kotzebue, since one of the early travellers explained to a correspondent that the original Otaheite rhymed with “ mighty.” It is still the pronunciation used by old residents and natives. t M. stands for George Matthai ; T.W.Y. for T. Wayland Vaughan ; Dod, for L. Doderlein and H.M.B. for H. M. Bernard. 90 GREAT BARRIER REEF EXPEDITION Other recorders from Gt. B. Reef. Wider distribution. Page. Plate. [Seriatoporidae] Seriatopora hystrix . T.W.V. . ? Red Sea ; Indo-Pacific to Fiji. . 107 . Stylophora pistillata T.W.V. . Red Sea, including Suez ; Indo- Pacific to Fanning Atoll. . . 107 . S. septata . Rotuma. . 108 . II Pocillopora damicornis Indian Ocean, including Natal and Mauritius ; Pacific, in- cluding Tahaiti and Hawaii. . 110 . P. danae * S. Indian Ocean, including Natal and Mauritius ; Fiji and Tahaiti. . 110 . P. verrucosa . Red Sea ; Indo-Pacific, includ- ing Marquesas and Tahaiti. . Ill . P. eydouxi T.W.Y. . Indian Ocean ; Pacific, includ- ing Marquesas, Tahaiti and Hawaii. . 112 . I [Astraeidae : with distinct calices and cyclic septa] Leptastrea purpurea T.W.Y. . Red Sea ; Indo-Pacific to Tahaiti and Hawaii. . 115 I, III L. ehrenbergana • Red Sea ; Indo-Pacific to Tahaiti. . 115 L. transversa Red Sea ; Indo-Pacific to Tahaiti. . 115 LIV L. bottae Red Sea to Hawaii. . 116 I, II L. pruinosa sp. n. . 116 III Cyphastrea chaleidicum • Northern Red Sea, Pacific to Philippines, ? Hawaii. . 117 C. serailia • Northern Red Sea to Philip- pines. . 118 C. microphthalma Northern Red Sea to Tahaiti. . 118 Echinopora lamellosa . Indo-Pacific to Fiji. . 119 E. horrida Malaysia to Fiji. . 120 IV [Oculinidae] Madrepora kauaiensis . Hawaii, Indian Ocean. . 121 Galaxea fascicularis . • Red Sea ; Indo-Pacific to Samoa. . 122 G. clavus . . i Quelch Indo-Pacific to Samoa. . 122 Acrohelia horrescens T.W.V. . Red Sea to Palao Is. and Fiji. . 123 [Montastreinae] Orbicella curta T.W.V. . Pacific to Tuamotu Atolls. . 124 0. vacua sp. n. . 124 II, III [Astraeidae : without cyclical septa] Faviafavus - Northern Red Sea ; Indo-Pacific to Samoa. . 125 F. favus var. crassidens var. n. . 126 XIV F. valenciennesi Red Sea and Indian Ocean. . 126 F. doreyensis . Indo-Pacific to Samoa. . 127 F. speciosa Northern Red Sea ; Pacific to Fiji and Fanning. . 127 5LADREP0RARIA, HYDROCORALLINAE, HELIOPORA AXD TUBIPORA— CROSSLAND 91 Other recorders from Gt. B. Reef. Wider distribution. Page. Plate. F. stelligera • Northern Red Sea ; Indo-Pacific . to Tahaiti. 128 • Favites halicora . T.W.Y. . Northern Red Sea; Indo-Pacific . to Samoa. 128 • F. abdita • * Northern Red Sea ; Indian . Ocean, including Mauritius and Natal ; Pacific to Fiji. 129 * F. viretis . T.W.V. . Northern Red Sea; Indo-Pacific . to Samoa. 130 • VI F. as p era Maldives, Japan, Philippines. 132 . V Goniastrea retiformis . T.W.Y. . Northern Red Sea; Indo-Pacific . to Samoa. 133 • G. pectinata . T.W.V. . Northern Red Sea ; Indo-Pacific. to Samoa. 135 • G. benhami G. mantonae sp. n. = G. K5 Kermadec Isl., Formosa. 136 • VIII and K18 S. Manton . . 136 . VII [Astraeidae : Meandroid species] Cynarina savignyi . Gulf of Suez. 137 IV Lithophyllia vitiensis Fiji and Loyalty Is. 138 IX Caulastrea simplex sp. n. . 139 III C. furcata M. G.B.R., China Sea, Tongatabu. 140 Acanthastrea echinata . T.W.V. . Northern Red Sea ; Indo-Pacific . to Tahaiti. 141 VIII, IX Lobophyllia corymbosa Quelch . Northern Red Sea ; Indo-Pacific to Tahaiti. 142 IX L. hemprichii . M. Northern Red Sea ; Indo-Pacific . to Tahaiti. 143 X, XXX Symphyll/ia recta M. Indo-Pacific to Samoa. 144 XI S. radians Oulophyllia crispa . M. Indo-Pacific to Tongatabu. Indo-Pacific to Bismarck Is. 146 147 Coeloria astraeiformis . T.W.V. . ? Red Sea, Indian Ocean, Fiji. 147 C. daedalea . M. ; . T.W.V. . Red Sea ; Indo-Pacific to Tua- . motus, but not Tahaiti. 148 XI, XII C. lamellina . T.W.V. . Northern Red Sea ; Indo-Pacific to Samoa. 149 Leptoria phrygia Hydnophora microconos . T.W.V. . N orthern Red Sea ; Indo-Pacific to Samoa. Northern Red Sea ; Indo-Pacific to Samoa. 150 151 [Merulinidae] Merulina ampliata . M. Northern Red Sea ; Indo-Pacific to Samoa. 151 Fungiidae Fungia echinata . Red Sea to Tahaiti. 152 F. actimformis Zanzibar, E. Indian Is. to . Rotuma. 152 F. scutaria • Red Sea to Hawaiian Is. ; Tahaiti. 152 F. paumotensis . Red Sea to Hawaii. 153 F. fungites . T.W.V. . Red Sea ; Indo-Pacific to . Marquesas. 153 92 GREAT BARRIER REEF EXPEDITION Other recorders from Gt. B. Reef. Wider distribution. F. cyclolites . . . Dod. F. patelliformis Stephanophyllia formosis- sima Herpolitha Umax Halomitra robusta . PolypJiyllia talpina* Podobacia Crustacea . [Pectiniidae] Oxypora lacera 0. aspera Tridacophyllia lactuca T. paeonia [Agariciidae] Pavona decussata T.W.Y. P. danai P. cactus P. varians P. duerdeni Pachyseris speciosa P. torresiana . Coeloseris mayeri T.W.Y. T.W.V. T.W.V. T.W.V. [Siderastreidae] Anomastrea irregularis [Thamnastreidae] Psammocora contigua P. exesa .... [Agathiphyllidae] Diploastrea heliopora* [Dendrophylliidae] Balanophyllia incisa, sp. n. B. yongei, sp. n. Dendrophyllia arbuscula . D. micranthus D. micranthus var. grandis, var. n. D. velata, sp. n. Turbinaria peltata . . H.M.B. T. peltata var. gibiari, var. H.M.B. n. T. bifrons T, frondens T. stephensoni, sp. n. T. mantonae, sp. n. . Red Sea ; Indo-Pacific to Philippines. . Malaysia to Samoa and Hawaii. . Kei Is., Philippines, E. Africa, Hawaii. . Red Sea to Tahaiti. . Malaysia and Pacific to Society Is. . Malaysia, as far as Santa Cruz. Ceylon, Maldives to Philippines. ' -^Doubtfully known. Indo-Pacific to Fiji. . Fiji and Philippines. . Red Sea ; Indo-Pacific to to Samoa. Northern Red Sea to Tahaiti. . Indo-Pacific to Tahaiti. . Northern Red Sea ; Indo-Pacific to Tahaiti and Hawaii. Hawaii. Northern Red Sea ; Indo-Pacific to Tahaiti. . G.B.R., Timor, Tahaiti. Indian Ocean (?), G.B.R., Philippines, Palao Is. . Equatorial E. Africa and Natal. 1 Red Sea ; Indian Ocean to Mauritius ; Pacific to Mar- quesas. . Maldives, Pacific to Fiji. . Gulf of Aden to Samoa. Doubtful. Red Sea. Torres Straits, W. Australia, Singapore, Fiji, Philippines. ? Indian Ocean, Malaysia to Fiji. West Australia. Fiji, ? Samoa. Page. 153 Plate. 151 154 XVI 154 155 155 156 . XII, XIII 158 159 160 160 161 161 162 162 . XIII, XIV 163 164 164 164 165 • 165 . XV, XVII 165 . XVI, XVII 166 166 XV 167 . XIV, XV 170 XIV 171 173 . LV, LVI 173 LV 174 . XVI, XVIII XIX 175 . XIV, XVI- XVIII, XXII . 176 XXI 176 . XXIII 178 . LV, LVI 179 . XVI, XVII, XIX * Not collected. — [A.K.T.] MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— C'ROSSLAXD 93 Other recorders from Gt. B. Reef. Wider distribution. Page. Plate. [Perforata] [Acroporidae] Astraeopora myriophthalma Northern Red Sea to Samoa . 180 and Tahaiti. Montipora granulosa H.M.B. . Macclesfield Bank. . 181 XXV, XXVII M. millepora, sp. n. . . 182 . XX, XXII M. prominula, sp. n. . 182 XXIV, XXVII M. digitata Banda, Solomons, Fiji. . 183 XXVI M. fruticosa . H.M.B. ; . T.W.V. G.B.R. only. . 184 . XXIII M. ramosa T.W.V. ; . H.M.B. (as M. indentata) Ceylon to Fiji. . 184 XXVI, XXVII M. fossae, sp. n. . 186 . XXVI M. foveolata Tongatabu and Fiji. . 188 M. socialis Madagascar. . 188 M. venosa var. angulosa T.W.V. . Red Sea ; Celebes, G.B.R., . 188 xxvi-xxvm Samoa. M. prolifera . Amboina, Carolines. . 190 M. undans, sp. n. . 190 XXIII. XXVIII M. sulcata, sp. n. . 191 XXVIII, XXIX M. tertia, sp. n. . 192 XXV, XXVIII M. verrucosa H.M.B. . G.B.R. to Hawaii, but not . 193 Samoa or Tahaiti. M. erythraea . Red Sea, Batavia. . 193 XXIV, XXVII M. foliosa Indo-Pacific to Palao Is. . 194 M. inf or mis H.M.B. . Cocos Keeling, N. Celebes, . 195 G.B.R. M. composita, sp. n. . 195 XXVIII, XXIX M. angularis , sp. n. . 197 XXV, XXVIII The 40 species of Acropora are here in alphabetic, not classificatory order. Acropora abrolanoides . Brook ; T.W.V. Singapore, G.B.R., Tahaiti. . 204 A. affinis Brook G.B.R., China Sea. . 205 . XXXIV A. armata Diego Garcia, Singapore, . 212 . XXXVII ? Tahaiti. A. aspera Brook G.B.R. to Fiji. . 205 . XXXIII A. brooki, sp. n. . 228 . XLI, XLVI A. bruggemanni Brook G.B.R., Singapore. . 221 A. canalis Philippines. . 219 A. cancellata . Louisiade Arch. . 225 XLI A. clavigera G.B.R. only ? . 226 . XL, XLI I A. corymbosa . . Brook Red Sea, Indo-Pacific to . 211 Tahaiti. A. digitifera . . Brook ; T.W.V. Madagascar, G.B.R. . 207 . XXXV A. elseyi . Brook G.B.R., Torres Straits. . 223 A. exilis . Brook G.B.R., China and Arafura . 223 . XXXIX Seas. A. fruticosa . Brook G.B.R., Samoa. . 220 A. gemmifera . Brook G.B.R., Fiji, Arafura Sea . 220 A. glochiclados Indian Ocean. . 213 . XXXIX A. grandis . Brook G.B.R. only. . 202 XXXI A. haimei . T.W.V. Red Sea; Indo-Pacific to Fiji. . 207 XXXIII, XXXV 1 GREAT BARRIER REEF EXPEDITION Other recorders from Gt. B. Beef. Wider distribution. Page. Plate. A. hebes Brook ; T.W.V. Malaysia, G.B.R. to Samoa. . 217 • A. humilis Brook G.B.R. and Fiji. . 221 A. hyacinthus* Brook Tizard Bank, G.B.R., Fiji. . 215 A. intermedia Maldives. . 200 . XXXII A. jeulini, sp. n. . 227 XLII A. laevis, sp. n. . 230 XLV A. latistella Brook G.B.R. and Samoa. . 216 A. lutkeni, sp. n. . 229 . XLI, XLVI A. macrostoma Mauritius. . 214 A. otteri, sp. n. . 229 .XLIII, XLIV A. pacifica China Sea, ? Samoa. . 202 XXXI, XXXII A. palifera Brook ; . Indo-Pacific ; G.B.R. to . 218 T.W.V. Samoa. A. patula Brook G.B.R. only. . 215 A. prolixa “ Ousima,” Samoa. . 226 A. pulchra Brook ; Keeling Atoll, G.B.R., China . 203 T.W.V. . Sea. A. quelchi Amboina, Solomons, Samoa. . 209 . XXXVI A. rosaria Brook G.B.R. to Samoa. . 224 XL A. spectabilis . G.B.R. only. . 220 A. squamosa . Brook ; . T.W.V. “ E. Indies,” G.B.R. . 216 A. surculosa . Brook Indo-Pacific to Tahaiti. . 214 . XXXVIII A. tubigera Singapore, Banda, Louisiades, . 208 Palao. A. variabilis var. pachy- Brook Red Sea, Indo-Pacific. . 222 . XXXVIII clados [Poritidae] Goniopora tenuidens T.W.V. . G.B.R., Amboina, Philippines. . 231 G. lobata . ? Red Sea, Amboina. . 232 . XLVII G. minor, nom. nov. H.M.B. . G.B.R. only. . 233 . XLVIII G. hirsuta, sp. n. . . 233 . XLVIII Pontes stephensoni, sp. n. . . 238 L P. Inaddoni T.W.V. . G.B.R., Pacific to Samoa. . 241 P. lobata Fiji, Fanning, Samoa, Hawaii. . 242 P. solida . Red Sea and probably whole . 242 Indo-Pacific. . 243 P. australiensis T.W.V. . G.B.R., Philippines. . 243 P. lanuginosa . Hawaii. . 243 LI P. evermanni . . Hawaii. . 243 P. densa T.W.V. . G.B.R. only. . 244 P. andrewsi T.W.V. . G.B.R. and Samoa. . 244 P. suppressa, sp. n. . . . 245 LI P. annae, nom. nov. H.M.B. . G.B.R. only. . 246 LII P. ( Synaraea ) vaughani . . 247 LIII sp. n. P. ( Synaraea ) hawaiensis Macclesfield Bank, Samoa, . 247 Hawaii. Alveopora irregularis, sp. n. . 234 . XLIX, L A. mortenseni, sp. n. . 235 XLIX Stylaraea punctata . Red Sea. . 236 L * Not collected. — [A.K.T.] MADREP0RAEII4, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA — CROSSLAND 95 Hydrocorallinae [Milleporixa] Other recorders from Gt. B. Reef. Wider distribution. Page. Plate. Millepora tortuosa . Not Red Sea, nor E. of Fiji. 249 LIII M. foveolata, sp. n. . Alcyonaria [Coexothecalia] 249 LIII Heliopora coerulea . Not Red Sea ; Indo-Pacific to Funafuti. 250 Tubipora musica Northern Red Sea, Indo-Pacific. but neither Samoa nor Tahaiti. 251 Corals are so extraordinarily variable, not only in growth form but in wliat one would think were fundamental structures, such as the numbers of cycles of septa, as well as in details such as those of septal teething, that species have been found to run together, in some genera, in a wholesale fashion. I feel that it is now necessary to be on one’s guard against applying results found in one genus to members of another without criticism or publication of proofs. One must be watchful for details, apparently trivial, such as in other groups are somid foundations for specific distinctions, and which, in some corals may be found constant. For example, as Vaughan shows repeatedly, septal toothing is often a vital distinction, while in other cases it is of no consequence, and may be due to mere lack of building power. (Crossland, 1931.) In particular, growth form varies so greatly with conditions that many “ species ” are later found to be merely the effects of, e.g., shallow water, sediment or surf. There has thus arisen a tendency to ignore form altogether as a specific distinction, action which, in many cases, may be premature. These ecologic variations have then limits for each species, and, until these limits are determined, it is not permissible to merge species on the ground that they differ in little but their growth form ; and the same applies to other variable characters. An extreme case is my new species Montipor a fossae (p. 186), which hardly differs from M. ramosa except in being of decidedly different form. These forms and others may later be proved, by experiment and observation on the reef, to be only ecologic variations, but, until that is done, they must be kept apart. It is much easier for future workers if species are kept distinct until proved to be variations than if they are confused now and have to be sorted out later. On the other hand I find, in this collection as elsewhere, cases of variation within one species which seem hardly believable, but which are clearly proved by the series examined. Perhaps I may be allowed to add that I make new species with great reluctance. In the following I have therefore given details of variation at the risk of being tedious, though using them for the creation of new names, or the resuscitation of old ones, as little as possible. My object is to give material upon which others may perhaps work, in cases where what we have is insufficient. How far are the colours of living corals specific ? Most of the evidence is of astonishing variations without specific or even varietal differences, e.g., the brilliant green discs of several Astrean species, the bright pink form of a species of Pocillopora, and local differences, such as the green-coloured species of Stylojihora described by Gardiner, the green species of 96 GREAT BARRIER REEE EXPEDITION Cyphastrea in this collection, all species being brown in the Red Sea. Among the Perforata the leafy Montipora of Tahaitian lagoons is usually violet or blue, but often may be pinkish or brown, while the Acropora cytherea nearby is distinguishable from the other stalked forms by its lilac colour. Turbinaria is a genus usually brightly coloured. T. mesenterina, in the Red Sea, is variable, but within the limits of yellow and yellow brown, while T. peltata of the Great Barrier Reef may be whitish or rose pink. More data are necessary, and it must not be assumed from the above that colour never has specific value, or varies without ecologic reason. A difference between the continental reefs and those of the far oceanic islands is that in the latter, as in the Red Sea, the tide rises only about 18 inches at springs, so that corals are very rarely laid bare, and the wonderful photographs of coral beds taken by Saville Kent, Sewell and this Expedition are not to be had from these seas. More important is the fact that the barrier edge is always under the influence of the surf, whereas where there is a tide of several feet the surf’s action is greatly lessened for the whole 24 hours over most of the month. Consequently the modifications in growth form induced by the surge in Tahaiti shown in the present writer’s papers (1928, pis ii and iii ; and 1939, pi. 12) are not found on the Great Barrier Reef to anything like the same extremes. As species-work is the handmaid of ecology, I give reference to the ecological papers of this series whenever possible, viz., Vol. Ill, No. 2, by T. A. Stephenson, Anne Stephenson, Geoffrey Tandy and Michael Spender, “ The Structure and Ecology of Low Isles and other Reefs,” 1931 ; and Vol. Ill, No. 10, S. Manton, “ Ecological Surveys of Coral Reefs,” 1935. Miss Manton’s scale drawings of parts of the reefs are particularly valuable ; count- ting “ heads ” alone may be deceptive, e.g., Mayor, on at least one occasion, found the number of “ heads ” greatest where the total growth was least. My labour has been much lightened by Prof. Matthai, not only through the simpli- fication made by his published works, but by his having named a large number of these specimens. I have referred to published illustrations wherever I have found those which agree well with the specimens of this collection, since in many species the name alone gives little idea of which is meant of a series of variations. The naming of species in the ecological and other reports was, naturally, somewhat provisional, and the following emendations are necessary : Names used in earlier reports of this series. Pocillopora bulbosa Galaxea musicalis Leptastrea agassizi Pavia clouei Pavia pallida Pavia pectinata Pavia acropora Pavia vasta Goniastrea solida Goniastrea K.5 and Goniastrea K.18 Fungia danai Psammacora gonagra for Doderleinia irregularis „ Dendrophyllia nigrescens ,, Names used in present report. Pocillojwra damicornis. Galaxea clavus. Leptastrea purpurea. Favia speciosa. Pavia doreyensis. Goniastrea pectinata. Favia stelligera. Favites virens. Goniastrea retiformis var. solida. Goniastrea mantonae sp. n. Not found, probably Fungia fungites, Psammocora contigua. Halomitra robusta. Dendrophyllia micranthus var. synonymous with 33 35 53 53 3 3 33 33 3 3 means synonymous with 3 3 33 for MADREPORARIA, HYT3ROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 97 Names used in earlier reports of this series. Names used in present report. Astreopora ocellata Montipora ramosa Montipora foliosa Acropora decipiens Acropora hebes Acropora delicatula Acropora formosa Acropora loripes Porites liaddoni probably synonymous with for includes for for Astreopora myriophthalma. Montipora fossae sp. n. (?) Several species. Acropora variabilis and others. Acropora intermedia. Acropora armata and (?) others, Not found in the collection. Acropora quelchi. Porites stephensoni sp. n. Acropora hyacinthns and Acropora cytherea are probably confused. Among differences from other coral regions, besides details noted under the species and genera, such as the absence, from the Murray Island Collection, as well as this, of the semi-encrusting species of Dendrophyllia, usually so abundant, it is notable that large colonies seem to be rare or absent. True, S. Manton (1935, p. 297) refers to “ huge ” Montiporas and Acroporas at 16 feet and below on the windward side of the Low Isles reef, but gives no idea of what “ huge ” means. Compared with the very small colonies S. Manton had dealt with in shallow water they may have deserved this description without being so very big. Were they over 12 feet across ? In all the coral seas familiar to me, and from descriptions in the literature, corals of this size are exceptional, but those of six feet in diameter are not uncommon in most coral seas, in the genera Pontes and Acropora especially, while Gardiner records cpiite large masses of Astreans, of species which are all quite small here, from many places in the Indo-Pacific. Eguchi (1938) records Montipora foliosa 3 meters across in Iwayama Bay, Palao Islands. What is the present effect of all this growth, the balance between deposition and destruction ? Paradice was the first (1925) to give any details of reef structure in this region, and described the remarkable columnar rocks found on the lee sides of many reefs, of which he gives a sketch on p. 54, which may be compared with the scale section by Stephenson and others (1931) on p. 66. Thanks to the diving helmet we now have under- water details which were invisible to Paradice, given by Stephenson and others (1931) on Plate XVII and by S. Manton (1935) on Plate XII. In both cases we see that the rock masses are smooth and bare,* and such coral as now grows thereon had no part in their formation, is adding nothing to the bulk of either reef edge or pinnacle, and does not even add anything but sand to the deposit at the base of the rock. It is clear that the lee reef edge is under erosion, and the pinnacles are remnants of once existing reefs now removed. To windward conditions are entirely different. Paradice, Stephenson and others all agree as to the steepness of the reef edge and abundance of coral growth, S. Manton especi- ally referring to the abundance of “huge” colonies, but unfortunately she was unable to see the character of the precipitous part of the slope, 28 feet deep out of a total of 86 feet. Such a steep sometimes indicates erosion, but, at any rate in its upper part, the reef is clearly growing * These edges are comparable to those of the Tahaitian lagoons, but there is much more coral below the level of Low Water Spring Tides here than in Tahaiti, where live coral was confined to the upper surfaces of the overhangs, which it did not, or only exceptionally, form. 98 GREAT BARRIER REEF EXPEDITION Now to take the reef as a whole we have first Paradice’s (1925) diagram on p. 53, which can be well compared with the aerial photograph* of Coates Reef given by Yonge in “ A Year on the Great Barrier Reef,” Plate XLI, opposite p. 137. (The diagrams and figures of Stephenson and others are essentially the same, but Coates Reef is the simpler case.) Both clearly show a solid growing edge of definite outline to windward, irregular growth overbalanced by destruction on the lee side. This definiteness and solidity are due to several factors : (1) More rapid coral growth, which is generally found to be accompanied by decreased activity of destructive organisms. (2) Growth of lithothamnionae, such as are abundant only among waves. (3) The mechanical action of the waves, which drives fragments into all spaces between the larger corals and so consolidates the reef. The reef is thus creeping to windward, like some of those in the Red Sea. (See Bertram, 1936, p. 1013). Imagine such a reef in open water, and subject to winds from two directions. The photograph and diagram may then be duplicated, and fitted together like a pair of horse- shoes with the open ends together — a miniature atoll. It is also noted that the barrier edge, at least among and just outside the breakers, though cut by trenches, is smooth and has no vertical step nor outlying buttresses, in fact just like that of Tahaiti. The expedition had no opportunity of seeing the reef just outside the breakers, but this conclusion is safely drawn from the photographs of regularly breaking waves. The same applies to many of the Maldivan and Pacific reefs from examination of photographs by Agassiz, Gardiner and Sewell, while some of those given by Agassiz show quite a different type of edge, abruptly falling, like those of Napuka Atoll, part of the exposed shore reef of Tahaiti, and of Funafuti . Judging by Agassiz’ photographs, 'part of the Funafuti reef edge is regular and smooth. The formation of platforms in the moat, and in general beds of living coral on reef flats and the tops of the Paradice pillars, raises an interesting question which has not hitherto been considered, viz., how can these living corals still exist seeing that the sea level has been nearly the same for the last few thousand years ? Why did not the whole surface become a platform of dead rock many years ago ? The answer evidently is that, in these areas, either the decay and removal of coral is going on as fast as its growth, or that there is an alternation of rock formation and decay in different areas, so that as fast as a platform appears in one place it is removed in another where coral growth is resumed. In short, the fate of the ivhole of the coraLliving on a reef flat is to be broken up into sand and mud, which is swept away into deeper water, where again, as Gardiner has shown, its fate is often further subdivision and finally solution. Permanent beds of living coral can exist only immediately along the edge of an outgrowing reef, where their foundations are being continually extended. Here follows for cross reference an index to the corals mentioned by Dr. S. M. Manton (Mrs. J. P. Harding) in Vol. Ill, No. 10, of this series. * The photograph of Coates Reef was first published by courtesy of the Royal Australian Navy as frontispiece to Prof. Richard’s paper, “ A geological reconnaissance in North Queensland,” Trans. Roy. geol. Soc. Australasia, I, 1924. — [A.K.T.] MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 99 INDEX TO NAMES OF CORALS USED BY Dr. MANTON. Sci. Rep. Gt. Barrier Reef Exped., Ill No. 10, 1935, Ecological Surveys of Coral Reefs. (In dassificatory order.) Roman numerals refer to Plates.* [Seriatoporidae] Seriatopora hystrix, 295, 298, 306, 307, VII, "\ III, X^ I. Stylophora, 298, 302, 306, 307. S. pistillata, 289, 295, IV, VII, VIII, XI, X^ I. Pocillopora, 302, XIII. P. verrucosa, 297, 304, 305, 306, VIII, XIV, XV, XVI. P. danae, 288, 298, 306, XII, XVI. P. eydouxi, 288, 297, 298, 305, 307, VIII, XV. P. bidbosa, 284, 288, 289, 292/3, 295, 298, 300, III, VII, VIII, IX, X, XVI. Astraeidae with distinct calices and cyclic septa.f Leptastrea, 284, 287, 288, 293, 294,^298, 300, 302, 305. L. ekrenbergana, 288, 298, 305, VIII, XII, XIV, XV. L. roissyana, 284, 287, 288, 293, 294, 300, 302, 305, III. Cyphastrea, 284, 295, 300, 302, III, VII. C. chalcidicum, 288, 293. Echinopora lamellosa E. sp., 295, XI, XII. Galaxea, 289, 295, 300, 302, XIII, XVI. G. musicalis, 284, 288, III, VII. G. fascicular is, 288, III, XI. Orbicella, XI. Astraeidae without cyclical septa. Favia (including Favites), 284, 287, 288, 289, 294, 300, 302, 304, 305. F. abdita, 285, 288, 294, 298, 305, 306, 307, VIII, IX, X, XII, XIV, XVI. F. acropora, 289. F. astraeiformis, 285, 286, 288, 292-4, VIII, IX, X, XI, XIII. F. clouei, 285, 288, 298, 302, 305, III, VIII, XII, XIV. F. doreyensis, 285, 288, 294, 297, VIII, IX. F.favus, 285, 286, 288, 293, 294, 298, 306, VIII, IX, XI, XII, XVI. F. halicora, 285, 288, 294, 298, 306, 307, VIII, IX, XI, XV, XVI. F. laxa, 305, XII, XIV. F. vasta, 288, 298, 302, 306, 307, III, VIII, XVI. F. wakayana, 305, XIV, XV. Goniastrea, 293, 295, 303, 305. G. pectinata, 284, 285, 287, 293, 294, 300, 302, III, X. * Dr. Manton’s Plates 1-VIII are diagrams of the traverses ; IX-XVI, pictorial diagrams of small areas. f It is difficult to believe that these “ species ” can really be recognized on the reef. [The following information has been supplied by Dr. S. M. Manton : “ Identifications of the corals encountered on the detailed surveys were not made on field observations alone. Both living and cleaned corals were examined in the laboratory, and full use was made of existing monographs. Corals which did not correspond with anything in the monographs were given provisional names. When an observer becomes sufficiently accustomed to examining the corals of a limited region it is found that many forms are clearly recognizable in the field. Those about which any doubt was felt were brought back to the laboratory for examination. It may be noteworthy that certain distinctions recorded in the field on living Xenias were completely invisible after preservation, but subsequent examination of the spicules by Professor J. S. Hickson confirmed that these superficially similar forms were quite distinct species. However, two corals which appeared in life to be unlike, “ K.18 ” and “ K.5”, are both referred to the same species, Goniastrea mantonae, by Dr. Crossland.” — A.K.T.] 100 GREAT BARRIER REEF EXPEDITION G. retijormis, 285, 286, 288, 293, 294, 297, 305, 306, 307, VIII, XI, XII, XIV, XVI. G. solida, 294. G. K.5, 288, 294, 305, 306, VIII, XI, XII, XIV, XV, XVI. G. K.18, 287, 293. Astraeidae Maeandrinae F. astraeiformis (see above). Coeloria, 284, 294, 295, 303, 305, VIII, IX, XIV, XV. C. daedalea, 288. Platygyra phrygia, 284, 288, 295, 298, 305, VIII, IX, XI, XII, XV. Lobophyllia, 289, 295, 298, 302, VII, VIII, XI, XII, XIII. Symphyllia, 284, 295, 298, 302, III, VII, IX, XI, XII, XIII. Hydnopliora, 298, 305, VIII, XIV, XV. Unclassified Merulina ampliata, 289, XI. Fungiidae, VII. Fungia, 284, 290, 302, 303, 306, 307, XIII, XVI. Herpetolitha, 303. Halomitra, XIII. Podobacia, XII. Polyphyllia, 290. [Thamnasteriidae] Psammocora, 301, 302. Ps. exesa, 295. Ps. gonagra, 284, 288, III, IX. Pavonx, XI. [Acroporidae] Montipora, 288, 289, 294, 298, 302, 303, 304, III, VIII, XII, XIV. M. foliosa, 293, 294, 295, 300, VII, IX, XI. M. erythraea, 294, VII, XI. M. ramosa, 284, 293, 294, 300, 301, 303, III, VII, IX, X, XIII. Astreopora, 290, 302, IV, VIII, XI, XII. Turbinaria, 289, 290, 295, 303, XII. Acropora, 289, 294, 300, 302, 303, IV, VII, IX, X. A. abrotanoides , 298, 302, 305, VIII. A. decipiens, 297, 298, 302, 305, 307, VIII, XIV, XV. A. delicatula, 295, 305, XV. A. exilis, 287, 295, 296, IV, VII', VIII, XI. A. formosa, 287, 295, VIII. A. gemmifera, 287, 295, 298, 303, 304, 305, 307, VIII, XI, XII, XIV, XV, XVI. A. hebes, 284, 293, 295, 300, 305, IV, VII, VIII, XI, XIV. A. hyacinthus, 287, 295, 298, 305, VIII, XI, XII, XIV, XV, XVI. A. loripes, 295. A. palifera, 305, 306, XV, XVI. A. polymorpha, 295, 296, 298, 303, IV, VII, VIII, XI, XII. A. pulchra, 284, 293, 295, 302, 305, VII, XIV. A. quelchi, 284, 287, 295, IV, VII, VIII, XI, XIV, XV, XVI. A. rosaria, 287, 295, 305, VII, XIV. A. squamosa, 284, 295, 297, IV, VII, VIII, XI. A. variabilis, 287, XI, XII. MADREPORARIA , HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 101 [Poritidae] Goxiopora, 289, XI, XYI. Porites, Massive, 284, 286, 288, 289, 290, 295, 300, 302, 305, III, VII, VIII, IX, X. XI. XII, XIII, XVI. Branched, 289, 302, VII. XI. [Dendrophylliidae] Dendrophyllia, XII. Tubipora, 304, XIV. Heliopora, 295, 298, VII, VIII. Millepora, 297, 298, 304, VIII, XI, XII, XIV, XV, XVI. [Family Caryophylliidae.] Sub-family Caryophylliinae . * My inclusion here of the genus Trachyphyllia is a consequence of my examination of the solitary corals of Prof. T. A. Stephenson’s South African Ecological Survey. It is based on the fact that in the genera Polycyathus, Paracyathus, and Caryophyllia the “ pali,” upon which classification is based, are parts of the septa and not independent structures. The three genera form a series which show this very clearly. In Polycyathus the septal teeth are of the same form from the top of the wall to the middle of the columella, as is particularly well seen in a section of the calyx. In the next stage, e.g., in Paracyathus, the lower parts of the septa are differentiated from the upper. This differentiation takes various forms : in Paracyathus the lower part is thickened and bears large, sometimes twisted teeth, the so-called pali ; in the adult Trachyphyllia it is entire, and broader than the slightly thinner upper and toothed part, but in young specimens its shortness suggests a palial lobe rather than a division of a septum. In this it resembles Caryophyllia, which has similar shortened lower portions, and in the common Mediterranean species, C. cyathus, this shortening has gone so far, and the notch between the two parts is so deep, as to excuse the idea of structures of independent origin. As this species was the first known, being- obtained in quantity by the fishers of Corallium rubrum, this appearance of distinct pali was stretched to cover Paracyathus and Polycyathus, whereas morphologically the series had its starting-point in the opposite direction. This division of the septa into two parts is found also in Favites virens (this report, p. 130), thus perhaps making a second connection with the Astraeidae. Genus Trachyphyllia. Trachyphyllia geoffroyi Aud. 1899. Antiilia sinuata Gardiner, p. 167, pi. xx, figs. 26a, 266. 1928. T. geoffroyi Mattliai, p. 95, pi. 22, figs. 1-11 ; pi. 23, figs. 1, 2, 5 ; pi. 26, fig. 1 ; pi. 60, fig. 1 ; pi. 62, figs. 1-3, 7, 8, 11. 1936. T. amarantum Yabe, Sugiyama and Eguclii, p. 22, pi. xii, figs. 1-12. * Crossland inserted in his MS. only two family and subfamily names. His descriptions of species were arranged in the same “ classificatory order ” as those in his “Index to Dr. Manton’s Corals” (p. 99), where he had inserted six headings. These and fifteen similar new ones have been inserted in the text in square brackets. It was thought best to retain the Author’s arrangement rather than to use the standard one of Vaughan and Wells (1943) published just after his death. — [A.K.T.] VI, 3. 8 102 GREAT BARRIER REEF EXPEDITION Yabe, Sugiyama and Eguchi also refer to Wells’ proposal to substitute Manicina for Tracfiyphyllia, and, quite rightly, reject it. All the specimens were preserved in alcohol, and most remain so. Stns. 22 and 23 provided two specimens each ; Stn. 19 gave 21 specimens, ranging from 15 mm. to 65 mm.; Stn. 16 one of 13 mm. All are top-shaped, and only one, of 62 mm. has a second small mouth. The deeply constricted and meandroid, large forms which occupy most plates of illustrations are not represented here, so that, of the 36 illustrations cited above, three of Matthai’s (pi. 22, figs. 2-4) and the two of Gardiner’s, with Savigny’s, are all that illustrate this series. Gardiner’s description and figures of his Antiilia sinuata correspond very exactly with several of these specimens, which are “ widely open, trumpet-shaped,” and may or may not be “ somewhat bent and twisted.” The epitheca is thin, as in Gardiner’s, in some specimens, thick and dense in others. In view of these facts I do not regard it as proved that there is but one species of this genus, Matthai’s series being, so far as his account goes, quite incomplete. There may be even three species, T. sinuata (Gard.) and T. constricta Briig. as well as T. geoffroyi, but sufficient data have not yet been published for certainty. If this turns out to be so the present species is T. sinuata (Gard.). The smallest specimen figured by Matthai (1928, pi. 22, fig. 1) is 45 mm. in longer diameter, a quite regular oval without constriction. The next to it is of the same size, but shows the beginning of a constriction. Neither shows the characteristic “ palial lobes,” and the calices appear shallow. In all but one of the Great Barrier specimens, even in the two of 12 and 15 mm., the palial lobes are conspicuous, clearly visible even while still covered by the flesh of the polyps, and the calices are deep, though comparatively shallow in some of those of medium size. Up to 37 mm. x 23 mm. only one specimen, of 30 mm., shows constriction, and that on only one side. The two smallest are circular, and there are none as large as Matthai’s fig. 1, which has not become elongated and begun constriction. I therefore doubt whether Matthai’s specimen is really the young of this species, since the present series is so complete. The smallest known specimen is in fact a Caryophyllia, from which the adult form differentiates by regular steps ; it is not a larval or ancestral form. In this smallest specimen 5 cycles of septa are developed regularly, of which the first two bear the palial lobes, and are thickened near the columella, which they form by sending out trabeculae. The third cycle rarely reaches it, and more rarely takes part in its formation ; the fourth reaches about two-thirds of the way down the wall, and the fifth is short on one half of the calyx, rudimentary over the rest. Specimens 25 mm. in longer diameter are the same, but the third cycle regularly joins the columella. In larger specimens septa multiply, and those of the lower orders are also broadened below, at first to a smaller extent, then to equal those of the first two orders. Distribution. — Northern Red Sea and Indian Ocean (rare in both) ; E. Australia and China seas, but no further east than the Philippines. Heterocyathus aequicostatus M. E. and H. 1904(a). Gardiner, Marine Investigations in South Africa, III, 1905. (This paper published in Sept., 1904.) 1909. Harrison and Poole,* p. 898, pi. 85, figs. la-/. Gardiner shows that there is only one species of this genus, the variation in which he * See footnote, p. 103. MADREPORARIA, HYDR OC'OR ALLTXAE , HELIOPORA AND TUBIPORA— C'R OSSLAXI) 103 investigated by detailed examination of nearly 900 specimens. His plate illustrates 43 of them. Harrison considers that two species can be distinguished. The Great Barrier specimens are only two, both young, from Linden Bank, Stn V. The larger is fairly closely represented by Gardiner's pi. iii, fig. 41, but the smaller, which has a disc measuring 6 nnn. x 4 mm., has a ridge-shaped base pointed at the closed end, apparently enclosing a very small shell completely. Distribution. — Gardiner's specimens were all from 40 to 90 fms., apparently all off the coast of Xatal ; Semper’s from the Philippines in from 6 to 25 fms. I have dredged it from about 10 fms. near Wasin, Equatorial East Africa, and it is recorded also from Cevlon, Burma and the Persian Gulf.* Rehberg's specimens were from West Australia and S. China. [Sub-family Thecocyathixae] Genus Thecocyatlms. Gardiner (1938, p. 173) places this genus in the Flabellidae, as the polyp has little or no edge-zone. Since the value of this feature in classification needs reinvestigation, and it is not known whether all the species of Flabellum are without edge zone,f and in view of the close likeness of Thecocyathus and Paracyathus in their septa, palial lobes and columella, I keep the genus in its old position for the present. Thecocyathus minor Gard. 1899. Gardiner, p. 163, pi. xix, figs. 3 a-b. I have three specimens of this species from a mass of rotten shells bound together by lithothamnioneae, obviously dredged (but no data given). The B.M. number is 414. The correspondence with Gardiner’s description is very close, but I am able to add notes on variation. The epitheca is thickly overgrown by nullipore almost to the edge of the calyx, so that no transverse or costal markings are visible. For !! extremely granular sides ” of the septa I should prefer to say septal sides with numerous spikes, which would become synap- ticula with a little more growth. These, with the black colour of the rest of the septa, which remains after boiling in soda, give this part of the coral a resemblance to Ulastrea crispata. As Gardiner found, the septal series are hard to see, and in my specimens the extra thickness and prominence of the primaries can only be made out sometimes. The “very irregular arrangement of its pali ” was noted by Gardiner as a peculiarity of the species, and this deserves further examination now that three specimens are available. The two divisions of the septal edges may or may not be clear, and are sometimes non- existent. The upper parts are nearly horizontal, and bear small blunt teeth, which increase in size and become flattened transversely to the septa on the nearly vertical part, and bear denticles. This enlargement amounts to great swellings in part of the circumference of one specimen, with or without a cut between them and the rest of the septum. In other * In the typescript Crossland had entered the title of the paper as “ Some corals of the Persian Gulf.” The corals were collected in the Mergui Archipelago. — [A.K.T.] f Fowler (1888) ; Moseley (1881), p. 164, pi. xvi, fig. 10. 104 GREAT BARRIER REEF EXPEDITION parts of this same calyx the knob-like teeth pass into the columellar processes, while in other septa the teeth remain small, in marked contrast to the large knobs of the columella. In another specimen the first enlarged teeth may be high up the septum, with one or two others rather smaller below it projecting well from the septum, pah- fashion. In the third these teeth, which are not much thickened, are separated by deep vertical cuts from the upper nearly horizontal part of the septum, and so appear pillar-like. There are usually two of these teeth, but occasionally one or three, and the distance of the first vertical cut from the wall varies greatly. Distribution. — Sandal Bay, 40 fms.; Lifu, Loyalty Is.; N.E. of New Caledonia, about 22° S. [Sub-family Eusmiliinae.] Euphyllia glabrescens (Chamisso). (Plate II, fig. 6.) 1899(a). E. glabrescens Cham. Gardiner, p. 735. 1904(6). E. glabrescens Gardiner, p. 759. 1911. E. laxa Gravier, p. 31, pi. 2, figs. 5-8. 1918. E. glabrescens (pars) Vaughan, p. 82, text-fig. 1, pi. 26. 1918. E. glabrescens Mayer, pi. 19. 1928. E. glabrescens Matthai, p. 174, pi. 42, fig. 5 ; pi. 44, fig. 4 ; pi. 62, fig. 9. One fragment, “ June Reef, outer Barrier, Nigger head, No. 431,” with three elongated calices and one nearly round. All the main septa being strongly exsert, generally 2 mm., and up to 2.5 mm., gives the specimen a very abnormal appearance, the level junction of the broad main septa with the delicate wall being one of the characteristic features of the genus. Of the 10 figures in which this point can be made out published by Vaughan, Matthai, Gravier, Faustino, Yabe, Sugiyama, Eguchi and Bedot, in seven there is no exsertion, in three a slight exsertion here and there. I was therefore inclined to make this specimen a separate variety until I saw Gardiner’s remarks on his Rotuman and Maldivan specimens. In 1899 (p. 736) he gives an account of variations of growth-form which disposes of Gravier’s E. laxa. “ The septa vary up to 2 mm. in exsertness ” “ The colour of the polyps varies from green- yellow to olive-brown, always markedly greener towards the stomodaea.” “ Rotuma ; in pools of outer reef, very local in distribution, often forming large masses, 3-4 feet across.” In 1904(6) (p. 759) he writes : “ The specimens differ from the Rotuma ones in that the septa very seldom rise above the wall.” The species “ may generally be found in hollows towards the inner side of the reef-flat or in protected situations, where there is no sand or mud. . . . Where it occurs . . .' it is exceedingly abundant, but is nowhere a reef-builder. Colour, dull green.” It may easily happen, especially in a species with so delicate a wall, that the wall does not keep pace with the upward growth of the septa, probably as a temporary growth stage. As, however, this striking and exceptional form has not been photographed I illustrate it in Plate II, fig. 6, a nearly side view to show also the rather prominent costae. I conclude that Low Isles is too muddy for this species. Distribution : Gulf of Aden, but not the Red Sea ; Indian Ocean and Pacific as far east as Samoa only. MADR EPOR ARIA, H YDR OCOR ALLIXAE , HELIOPORA AND TUBIPORA— CROSSLAND 105 (?) Euphyllia fimbriata (Spengler). A very small fragment from Batt Reef, B.M. 268, is all that represents this species, but its valley is only 8 mm. across. The fragment is too small to be worth further discussion. Distribution : E . fimbriata has already been recorded six times from E. Australia. It occurs in the central Red Sea,* and there are numerous records from the Indo-Pacific, but it has not been found even so far east as Fiji. Family Flabellidae. The relationship between the genera Flabellum and Euphyllia is obviously close, so much so that Dana placed two species of Flabellum in the genus Euphyllia. f In 1904 Gardiner (p. 954, pi. 93, figs. 28, 29) described an elongated Flabellum with several stomo- daea, F . multifore, which, in top view, resembles E. fimbriata so strikingly that Matthai (1928, p. 179) puts (?) F. multifore in the synonymy of E. fimbriata. There is in the Koben- havn museum a specimen of a PlacotrocJius of about the same size as F. multifore which I hope to describe later, when this whole subject can be thoroughly discussed. Possibly Coeloseris Vaughan (or Aplocoenia M. E. and H.) may belong to this group. There has been much discussion about, and great importance attached to, the structure of the wall in this family. It is described as “ epithecal,” a very unsuitable term as the word means added upon the theca, whereas, as there is no edge zone in Flabellum, % the wall is deposited from the inside only, not from both sides as in most corals. It is therefore better described as a hemitheca. Two entirely distinct things have been confused under this name epitheca, apparently only because both result in dead-looking surfaces, (1) the hemitheca of the Flabellidae, § and (2) the covering of the lower part of the skeletons in many corals, especially conspicuous in, e.g., Trachyphyllia (or Antiilia as it is called in the literature of this subject.) This is a generally dull-surfaced deposit upon a theca which has been made in the usual double way, and so is a real epitheca. It is secreted by the lower border of the edge zone in its retreat as the coral grows upwards. || Von Koch (1886) is responsible for this confusion, the lack of logic in which was exposed by Fowler (1887 p. 15.) Von Koch gave importance to “ dark lines ” in the skeleton, but apparently knew nothing of the living polyp. It is von Koch, however, who has gained the attention of his successors. Genus Flabellum,. Flabellum rubrum (Q. and G.). 1904. Gardiner, p. 125, pi. iv, figs. 22-34. One specimen is present, dredged from Stn. XIV, 19 fms. near Lizard Island and named * This is the first record from the Red Sea. I especially noticed it at Dongonab, being attracted by the difference between its polyps and those of Lobophyllia. The living colonies resemble each other, but the polyps of the former show their tentacles by day (or at least in the early morning) when those of the latter are retracted. t Yonge (1930, p. 20) shows that Euphyllia has an edge-zone. J See Yonge’s figure (1930, text-fig. 1) of the expanded polyp of F. rubrum. §According to Fowler (1885, p. 586) F. patagonicum may have a retractile edge-zone and so a complete theca. F. pavoninum also has a smooth porcellaneous outer surface, but nothing is known of the living polyps of either form. || There is evidence to suggest that both hemi- and epitheca are physiological responses to conditions e.g., sand rasp and burial, or the growth of organisms round the bases of fixed forms. 106 GREAT BARRIER REEF EXPEDITION by Prof. Matthai. It resembles Gardiner’s fig. 31, pi. iv, but has a third root, very short, 1 cm. up one side, and the coral is broken off at a level between the two longer roots. Distribution : Indian Ocean, Maldives and south to Cape Colony. Of the eleven localities given by Gardiner, eight are east and north-east of Cape Agulhas, but four are not to be found on an ordinary good atlas ; of the seven known localities five are off Natal, and from these 491 of the 584 specimens were obtained. Known also from the Philippines and Palao Islands in the Pacific, and as far south as New Zealand and the Bass Straits. Not recorded east of the Palao Is. Flabellum vacuum, sp. n. (Plate I, figs. 1, 3.) Three specimens, dredged Stn. XVI. mm. mm. mm. Calices 45 x 28 ; 24 x 16 ; 13 x ? (broken but nearly circular). Heights 32 22 15 The distinctive features of the species are : Broadly oval calyx, nearly round in the youngest, with narrow septa, leaving the cup almost empty. Columella absent in young specimens, in the largest rudimentary, formed mostly of septal teeth. Whole corallum is light and delicate. There are 24 principal septa, between each pair of which are one smaller, running down the wall to the columella, one very small reaching but a small distance down the wall, and from two to six represented only by short ridges at the top of the wall. That is to say, there are three cycles of nearly equal septa, the fourth smaller but complete, the fifth rudimentary and the sixth both rudimentary and incomplete. In the largest specimen* the costae of the first four cycles, with traces of the fifth, run down the outside of the calyx to its base as very low but distinct ridges, but they are only visible under a lens in the two smaller. The septa and costae project above the edge of the wall at most 1 mm., but, as always in this genus, this projection carries a portion of the wall with it, so differing from the exsertion of the septa of other corals. In the smaller specimens the fifth cycle of septa is often, and the sixth is always, absent. The species comes nearest to F. stabile Marenzeller, (1904, p. 273, taf. xvii, figs. 12a, 126), but this is an Atlantic deep-water form, and, among other differences, much coarser in structure. [Family Seriatoporidae.J Genus Seriatopora. Stephenson and others (1931) refer to the genus on pp. 67, 86 and 88. Characteristic of seaward slopes and anchorage at Low Isles, of Yonge Reef and the reef patch off Lizard Island, i.e., wherever water is clear. S. Manton refers to the genus on pp. 295, 298, 306 and 307, and on pis. vii, viii and xvi. It occurs on seaward slope of Traverse II beyond 364 feet ; on the windward slope in the third 100 feet, anchorage of Yonge Reef. PI. xvi shows some fair-sized colonies, * Selected as Holotype. — [A.K.T.] MADREPORARIA, HYDBOCOBALLTNAE, HELIOPORA AND TUBIPORA— CROSSLAND 107 the largest 20 cm. in diameter, in contrast to the exceptionally small colonies brought home (7 cm. in diameter, and one colony 13 cm. long but of stunted growth). Seriatopora Jiystrix Dana. Samples 37, 281, 195, 196 and 353 have been named by Matthai. These, like No. 440, are small but of normal growth. The spirit specimens are branches of larger colonies with crab galls. B.M. 276, 13 cm. in longer diameter, is abnormally flattened in shape, with frequent dichotomies and very short end-twigs. B.M. 388, “ stout pink Seriatopora nigger head, Undine Reef Y,” is comparable to Marenzeller’ s pi. 29, fig. 115, which he shows to be a thickened pathologic form. Distribution : Recorded from Murray Island by Vaughan and common through the Indo-Pacific as far as Fiji ; but the genus does not extend to Samoa, Tahaiti, or Hawaii. It may be the same as S. angulata Klz. of the Red Sea, but the material present is not sufficient for settling the question ; and, though Marenzeller has made a very full examination of this species, his figures are spoiled by being so roughly printed. Genus Stylophora. Only six small pieces of this genus have been brought home, one of which may be a nearly complete, though very small colony. Stephenson and others (1931) refer to the genus on pp. 86 and 88 as characteristic of seaward slopes and anchorage, and of Yonge Reef and Lizard Island, in all three cases accompanied by Seriatopora and Poeillopora. S. Manton refers to the genus aDd to S. pistillata on pp. 289, 295, 298, 302, 307, and plates iv, vii, viii, xi, xvi, on the steeper part of the seaward slope of Traverse I small colonies of 3 to 6 inches, which, like Goniopora and Turbinaria , are able to form normally shaped colonies in spite of the mud. Also on the seaward slope of Traverse II and the third 100 feet of Traverse III, but in shallower water on the windward side and accom- panied by Poeillopora and Seriatopora. It appears to be one of the less abundant corals in this region, and the colonies are generally small. The genus is distributed from the Red Sea (where it is abundant) to Murray and Fanning Islands and Fiji. It is not recorded from Samoa or Hawaii, and in Tahaiti occurs only as small thin crusts resembling Klunzinger’s (1879) S. armata (taf. viii, fig. 12). While it seems to accompany Poeillopora in local distribution it is absent from the regions in which the latter especially flourishes. Stylophora pistillata (Esper). S. digitata Klz., Marenzeller, 1906, p. 77. Marenzeller gives no reasons for considering this species identical with digitata and most of Klunzinger’s species, but the chances are that he is right. Specimen No. 2 consists of nearly straight, roughly cylindrical branches, and, as is often the case in the others the columella is not stylar. No. 311 (Lizard Island, A reef) is bushy, and with more flattened branches than Nos. 3 and 1, but is clearly the same species, probably grown in rougher water. The growth forms due to varied conditions shown by Klunzinger and Marenzeller are, in short, here represented. The best illustrations of the normal form of this species are those given by Savigny 108 GREAT BARRIER REEF EXPEDITION on pi. 4 (Polypes), figs. 3 (1) to 3 (5), Klunzinger coming second, but his specimens are somewhat stunted, as he collected at Qoseir in roughish water. Distribution : The species is abundant throughout the Red Sea, occurring even off the muddy beach at Suez, the Indian Ocean, Pacific to Fanning Island, but not to Tahaiti or Hawaii. Stylophora septata Gardiner. (Plate II, fig 5.) Gardiner, 1898, p. 996, pi. lxii, fig. 1. I refer specimen 369 to this species with some hesitation ; I have seen nothing else like it in nature or the literature. Unfortunately the specimen is but a fragment, measuring 75 mm. across and 60 mm. high. It has the growth form usual in specimens of several species grown in rough water, as were all four on Gardiner’s plate lxii. The weight of this example strikes one as unusual as soon as it is handled. To the naked eye the closeness of the calices and their large broad hoods are conspicuous. Even on the upper ends of the branches there are stout walls between the calices, perhaps more so than in Gardiner’s specimen, but this crowding of calices and thinning of their walls and coenenchyma always varies greatly with con- ditions of growth. The obtuse broad lips projecting 1 mm. over the mouths of the calices, often so developed as to resemble a downwardly directed tube, with striae outside so continuous and prominent as to resemble costae, are remarkable features. (Plate II, fig. 5.) The coenenchyme between is covered with scattered small pointed spines. The calices of the ends of the branches are deep, and open, the primary septa being thin, secondaries rudimentary, and the columellar style low down in the calyx. The top of one branch is intermediate between these conditions and those of the calices of the sides. These, the normal calices of the colony, are remarkable for the thickness and roughness of the primary septa, usually as thick as the spaces between them are broad, often thicker ; they reach nearly to the centre of the calyx, but the columellar style is, more often than not, invisible, though it sometimes appears as a small point near the surface. Secondary septa are present and distinct, but much shorter and thinner than the primary. This specimen came from the anchorage zone of June Reef ; Gardiner’s from Rotuma, outer reef. Gardiner remarks, “ This species is evidently very closely allied to S. digitata, having almost precisely the same mode of growth, and may perhaps be only a variety of it due to a very slow growth owing to its position on the reef, or some other cause. However, the presence of 12 distinct septa in nearly all the calices and the very obtuse lip are constant features of difference.” , I should add “ long ” to “ obtuse,” and observe that the growth form is common to all species grown in rough water. If this were merely a rough water form of S. pistillata (S. digitata) it would be similar to the S. mordax Dana described by Vaughan (1918), p. 81, pi. 25, and by him considered as probably a growth form of this species, but this is quite distinct from S. septata. Genus Pocillopora. I have made notes on the Family Seriatoporidae and the genus Pocillipora in my account of the corals of the Natal Coast (1948, pp. 179, 183). The results are briefly as follows : (1) The distinctions, between the genera PociUopora, Stylophora and Seriatopora, MADREPORARLA HYDROCORALLIXAE, HELIOPORA AND TUBIPOR A— CROSSLAND 109 though so clear in most cases, all have connections through species which show one or other of the characteristic features. (2) The variation in the species, especially of Pocillo- pora, is enormous (see, e.g., Vaughan, 1918, pi. 21) and the clue to it has not yet been discovered, nor can be without observation and experiment on the reef, as well as in a marine laboratory. (3) It is probable that such work would bring down the number of species in this genus to about six, but if Vaughan’s work on P. caespitosa — bulbosa — damicornis, etc., applies to other species it might be even smaller. Gardiner wrote in 1898 (p. 942), " The complete absence of verrucae on the tops of the branches and their incrassate (i.e., thickened) form in such a species as P. grandis [i.e., P. eydouxi] are due to the ends of the branches having reached the low-tide level, and, being unable to grow further upwards, increasing in both thickness and breadth. Although this is by far the most common species of Madreporaria on the reef at Funafuti, I never found any of its branches with their summits dead, even though they reach almost invariably to the low- tide level.” This is exactly what I found in Tahaiti, except that I should write “ genus ” for ££ species of Madreporaria.” I conclude, therefore, that the presence or absence of verrucae, close crowded thin-walled calices or those of more normal type on the ends of the branches, as well as the exact shape of the branches, are not of specific value, so that I am confirmed in thinking, e.g., Hoffmeister's P. setchelli from Tahaiti (1929, p. 359), of which he gives three figures, of growth form only, to be only a surf-flood variety of some other species such as P. verrucosa. Gardiner further gives the colour of the living colonies as “ usually green or pink when the polyps are expanded, but if retracted are nearly colourless.” Apparently there is an extraordinary variation in colours with locality, as in some other corals. I have seen a green Pocillopora only once, a very dark green, in the southern Marquesas. In these islands, and Tahaiti and the Red Sea, the colour, polyps expanded or not, is, with one exception, brown ; in P. eydouxi dark almost chocolate, ends of branches pink or white ; a yellower and fighter brown, with yellowish ends of the branches in P. verrucosa ; and a quite light yellow brown in P. damicornis. The exception is that P. verrucosa is sometimes bright red, the reason for which is unknown. I give a table of the common Indo-Pacific species, but, as I said, we have not the data for a useful discussion of synonymy. Gardiner wrote (1897, p. 942) : “ In the collections there are over 50 specimens . . . and in addition I have examined a very large number in the British Museum . . . and I am doubtful whether all these so-called species should not rather be described as varieties of one species, the characters of which would be the characters of the whole genus.” Table of Common Indo-Pacific Species of Pocillopora. T. W. Vaughan (1907) p. 100 (1918) p. 78, writes “P. damicornis, danae, verrucosa, meandrina, and elegans form a series so indistinctly broken that one is led to suspect that they are really continuous. It is probable that P. brevicornis and P. lobifera are a part of the same series.” A. Septa and columella rudimentary or obsolete (except on basal expansions, and in the varieties laysanensis and stylophoroides of P. cespitosa (Vaughan, 1907, pp. 88, 89)). (1) Small clumps, not exceeding 15 cm. in diameter : cespitosa — bulbosa — damicornis ; almost certainly one species. (2) Clumps exceed 18 cm., septa as spines, no columella, branches subterete : P. danae. Branches flat : P. verrucosa, P. meandrina, P. elegans ; all very close, if not identical. B. Septa and columella well developed, large clumps, big fiat branches (almost always) ; P. eydouxi and P. woodjonesi. These two species differ in ornamentation of spicules and details of columella and septa. 110 GREAT BARRIER REEF EXPEDITION I much regret that I can obtain no specimens from Ghardaqa which might help to decide whether Klunzinger’s species, favosa and hempriehi are the same as danae — dami- cornis — verrucosa, as seems very likely. This would give a distribution of these species from the extremes of the Indo-Pacific, from the Red Sea to the Marquesas and Hawaii. P. eydouxi I have not seen in the northern Red Sea. Pocillopora damicornis (Pallas) Dana. The names damicornis, cespitosa, bulbosa and acuta are discussed at length by Vaughan (1918, p. 75) and by Hoffmeister (1925, p. 15). Stephenson and others (1931) record the species P. damicornis from a series of habitats where coral growth is unexpected and much restricted, for example the sandy pool in the mangrove swamp, in the mangrove park, sometimes on mangrove roots and on the reef flat. It is one of the most abundant in the moat and the barren sandy flat of Batt Reef, “ low massive species ” being found near the edge. In Tahaiti it is common close inshore, even right at the foot of the beach at Pa’ea. It is this species which is generally found as young colonies, e.g., on floating logs (Wood Jones), chains of buoys, a wire-fence, pearl oysters, and tiles (Crossland, 1928), and a coco-nut (in the Kobenhavn museum) ; other coral larvae settling much more rarely, if ever, on artificial substrata. S. Manton distinguishes between P. damicornis (called bulbosa in her paper) and P. danae, probably including all the thicker-stemmed varieties of the former in the latter species if the two are really distinct. Her references bring out the fact that the thin- stemmed P. damicornis is “ extremely resistant to exposure, and to a wide range of conditions,” as we have learned from Stephenson and others. She also shows how, when conditions improve, e.g., in deeper water, or on windward or outer reefs, the change to P. danae or P. verrucosa takes place. P. eydouxi is less accommodating, being confined to outer regions and windward slopes, and corresponding with a more distinct specific differentiation. The following 8 of the 10 specimens have been named damicornis by Matthai : No. Ill, “ Prob. moat form ” ; No. 446 ; No. 21 ; “ P. 51 ” ; “ P. 40,” ££ Batt Reef, wave zone ” ; No. 324 ; No. 374 ; ££ June Reef, Int. Mad. Zone ” ; No. 325, ££ Lizard Island, A Reef” ; the other two by me, viz., No. 386, ££ Outer Barrier June Reef, Nigger Head ” ; and (no number), ££ Low Island, ordinary moat Pocillopora. ” Of these No. Ill and (no number) are good examples of the bulbosa form ; ££ P. 40 ” and No. 325 are thick, branched and grade into P. danae ; Nos. 374 and 324 are small, about 6 cm. diameter, with close short branches, obviously stunted ; while the Batt Reef wave zone fragments, ££ P. 40,” show no sign of stunting. No two specimens are alike ; the collection is an excellent series of the variation of this strange species, but, as all have been already illus- trated, especially by T. W. Vaughan in 1907 and 1918, there is no need to publish more. Distribution : The species is abundant from Singapore to Fanning Island in the Central Pacific, Tahaiti, and Hawaii. From the Indian Ocean Vaughan records it from Cocos-Keeling, but Prof. Stephenson’s collections enable me to add Natal and Mauritius. It is not recorded from Equatorial East Africa or the Red Sea, unless Klunzinger’s P. favosa is an extreme variation. Pocillopora danae Verrill. S. Manton refers to this species on pp. 288, 298, 306, and on pis. viii, xiv, xv, xvi. It is hardly possible to distinguish P. danae from thick specimens of P. damicornis on MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIRORA— CROSSLAND 111 the reef. For instance the reference on p. 288 might be read, “ Thin stemmed P. dami- cornis occurs in scattered colonies (2 to 12 inches), but beyond 420 feet in the deeper water the branches thicken" ; or on p. 298 on the seaward slope of Traverse III might read “as in the deeper waters of Traverse 1 on the lee side P. damicornis here grows with thicker branches, like those of P. danae.” Only one fragment, 8 cm. in diameter x 6 cm. high, is thus labelled by Matthai. It corresponds better with Vaughan’s (1918, pi. 22, fig. 2) Verrill’s type of the species, than with the figures of his specimen from Murray Island. It corresponds well also with Hoffmeister’s (1929, pi. 2, fig. 2) a specimen from Tahaiti. Specimens 325 and “ P. 40," both labelled P. damicornis by Matthai, an identification with which I agree, form an interesting series with P. danae. The first is a rather coarsely branched P. damicornis without any doubt, much like Vaughan’s (1918, pi. 21, fig. 2), while “ P. 40,” being from the wave zone of Batt Keef, is partly the same, partly with thicker branches not distinguishable from those of P. danae. I therefore think it very likely that this species is, in fact, a variety of P. damicornis , as does Vaughan, but I cannot dogmatize without further evidence. The species is recorded also by Thiel from Banda, but not by Umbgrove from Java or by Yabe, Sugiyama and Eguchi from Japanese seas. Faustino (1927) mentions it from the Philippines, but his figure (pi. 13, fig. 3) is the same as that already given by Vaughan (1918) in pi. 22, fig. la. Gardiner collected it from Funafuti and Fiji, but gives no figure. It is thus found throughout the Pacific as far east as Tahaiti. It has not hitherto been recorded from the Indian Ocean, but two of the specimens of P. damicornis in Prof. T. A. Stephenson’s collection from Natal are much like the intermediates mentioned above. In the Universitetets zoologisk Museum, Ivobenhavn there are 6 good-sized and com- plete specimens from Mauritius (labelled P.f rondosa, n. sp. by Lfitken, but no description was published) which clearly belong to P. danae, and resemble Vaughan’s figure of Verrill’s type, with rather long irregular verrucae, often a little swollen at then ends. The growth forms are an interesting series, from very loose and irregular to compact. I hope to publish illustrations in the future. Prof. T. A. Stephenson’s collection from Mauritius also contains a small specimen (No. 206) clearly belonging to this species or variety. We thus extend its distribution from the southern Indian Ocean as far as Fiji and Tahaiti. It is not recorded from the Red Sea or tropical East Africa, and figures of P. favosa KLz. do not resemble it. Pocillopora verrucosa (Ell. and Sol.). Manton, 1935, pp. 297, 304, 305, 306, pis. viii, xiv, xv, xvi. Manton (1935) says: “On the first 100 feet of the Seaward Slope [Traverse III to windward, corals] resemble those present on the leeward side . . . with the addition of large and small massive species of Pocillopora (P. verrucosa and P. eydouxi) and Acropora decipiens. These corals do not occur elsewhere on Low Isles, and are characteristic of the exposed outer ridge and reef crest of outer barrier reefs such as Yonge Reef.” This corresponds with my own experience, but pis. viii to xvi indicate that it is nowhere as abundant as in Tahaiti, though recorded as frequent on Yonge Reef. The collection consists of one specimen and two fragments, all named by Matthai, 112 GREAT BARRIER REEF EXPEDITION viz., 332 Outer Barrier, Ribbon Reef, Inner moat ; 342 from tbe same, but the outer moat ; and 74, a large branch, dividing dichotomously so that the upper surface has 24 endings. Also a small scrap, 285, with one of another species too small for identifica- tion. The specimens are typical, and call for no comment. Distribution : Red Sea ; probably Mauritius ; Cocos Keeling in the Indian Ocean ; throughout the Pacific to the Marquesas and Hawaii ; represented in the latter by “ the closely related, if not identical, P. meandrina Dana ” (Vaughan, 1918, p. 78). Pocillopora eydouxi M. E. and H. (Plate I, fig. 2.) The synonymy is given by Vaughan (1918) and Umbgrove (1939). Note Umbgrove’s remarks on Thiel’s new species P. symmetrica, an example of founding a new species on one broken scrap, which, in all corals, but especially in such a genus as this, is a proceeding which should rarely be attempted. S. Manton refers to the species on pp. 297, 298, 305 and 307, and on pis. viii and xv. The species is characteristic of clear water on exposed reefs.* It is, in almost all cases, recognizable at once on the reefs by its large size and flat branches, the only species with which it might be confused being P. elegans, grandis and elongata, three species of Dana which Gardiner gives as synonyms, and P. ivoodjonesi Vaughan, which is well distinguished by its large frosted spicules and inconspicuous columella. The colour of P. eydouxi is a cool brown with the ends of the branches pinkish or white. Its broad-ended branches and comparatively cellular structure make it possible for storms to break up the colonies, but, at the same time, the coarse gravel thus produced is very resistant, so that it is a particularly important reef and beach builder. I have published (1928a, p. 582) photographs of reef islets composed entirely or largely of this species. The present collection contains three specimens and four fragments, all but one named by Matthai (this one is quite typical). Nos. 269, 367, 268, and 274 are from June Reef, of the outer barrier series ; 402 and “ P. 50 ” from Low Isles, the former from out- side the rampart ; the latter, from the seaward slope, is the “ tip of a long flattened branch similar to those of specimen 364.” There is no note for 364. Three very distinct forms are present : A. The normal form with broad-ended branches, of which there are the three fragments “P. 50, ” 402 and 367. They are the triangular tops of branches : the last is prolonged below into a flattened but narrow stem, 20 mm. x 27 mm. None of them reach sufficiently near the base to show the typical calices with thick septa of Vaughan’s figure. From the literature, and from my own observations in Tahaiti, this is the normal and abundant form ; perhaps this is why these few scraps were considered sufficient to represent it ? The other specimens, much larger, are rare varieties, and, I suppose, were specially selected. B. Two large parts of colonies, Nos. 268 and 274, both from June Reef, con- sisting of a number of nearly cyhndrical branches, springing from common * Wood Jones (1907 ; 1910) says, “ Inhabits only tbe still deep water of tbe lagoon.” Tbis is contrary to Gardiner’s, Manton’s Umbgrove’s and my experience, in four widely separated regions. MADREP0RARL4. HYDROCORALLTXAE, HELIOPORA AND TUBIPORA— CROSSLAND 113 stems, the whole measuring 30 cm. high, by a maximum diameter of 21 cm., and 27 cm. x 21 cm., there is also an isolated branch 16 cm. long with 4 branchlets above. This form has been recorded only by Yabe, Sugiyama and Eguchi (1936), pis. ii, iv, fig. 2, v and vii, fig. 6, from the Marshall Islands. These three specimens differ from the others of the Barrier collections and from that of Yabe, Sugiyama and Eguchi in the irregularity of their verrucae, which in all three show a tendency to grow out into branches, the beginnings of such a curious growth as that shown by Gardiner (1897) in pi. lvii, fig. 3, under the name P. grandis. C. No. 269, of which the base is unfortunately missing, consists of five broad thick roughly triangular lobes, connected by broad bases, as shown on Plate I. The largest, which is incomplete, is 13 cm. wide, and 18 cm. high. The next largest, which is complete, is 15-5 cm. wide and 10 cm. high. Thickness of the largest, across the broken edge, 6 mm. at the top, 18 mm. near the base, the common base of the whole bemg 32 mm. thick. All surfaces, except the upper edges, are covered with rather closely placed smallish verrucae, remarkable for their regularity of shape and distribution. I have carefully cleaned the specimens, and examined the calicinal characters and spicules, as a result of which I endorse Matthai’s identification of these three remarkably different forms with the single species P. eydouxi. At the same time some ecological information about them is much to be desired, since calicinal characters alone are not certainly specific in this genus. The variation of these characters has not yet been described, though referred to by Umbgrove. The few figures given, by Milne Edwards and Haime, Vaughan, and Yabe, Sugiyama and Eguchi, all show the calices on the flat surfaces between the verrucae, and I may say at once that these are far more variable than those on the verrucae themselves, which are, however, owing to their position and delicacy, difficult to photograph. Such regular and thick septa and columella as Vaughan shows (1918, pi. 24, fig. 2a) are only found on the lower part of the branches, 15 cm. at least below the ends, higher up they become more dehcate and more variable, the septa varying in both breadth and number. As this collection contains only upper ends of branches in most cases, these typical calices are present only in the two larger specimens Nos. 268 and 274, i.e., those with cylindrical branches. Thus, the two short branches of set A have, on the flat surfaces septa 0 to 6 or 8 in number, narrow and thin ; near the upper edge generally only 0 or 2 or 3, or calices com- pletely empty ; columella either detached or joined to one or a pair of directive septa. In B.M. 402 [no Expedition number] columella is small and detached, often only traces of it visible. In No. 367, which is 14 cm. long, triangular end part 7 cm. long by 9-5 cm. broad, calices of the lower part of the stem like Vaughan’s figure, but septa and columella are delicate, and columellae all detached. Half-way up the stem the septa appear as two series by the narrowing of half of them, and one or a pair of directives are developed, but they do not unite with the columellae. On the upper triangular part calices very degenerate and shallow, and much of the coenenchyma is smooth. On the other hand, the calices of the verrucae almost always have 10 to 12 septa 114 GREAT BARRIER REEF EXPEDITION clearly seen, and the columella is detached, or joins a single directive, except close to the top of the branch where the septa are mere lines of little spines and the columella is absent. In No. 367 there are 12 septa and detached columella everywhere, but calices very delicate near the top of the branch. Form B with cylindrical branches : No. 364, a single branch 16 cm. long. Irregular verrucae, as in the two larger specimens. Near base, 6 very small delicate septa and prominent columella, but septa often absent, or nearly so. Columella may broaden, and join the septum on its upper side. In the middle the same, but a complete cycle of septa is rare, and columella usually attached to one directive. On the upper part of the stem are 12 septa, columella attached to one or two directives, or these may come close to the columella without attachment. No. 268, 21 cm. long, near base 6 to 12 rather delicate septa, in middle 6 to 12 as above sometimes very distinct, sometimes almost absent ; columella also absent sometimes. In the two large specimens the basal parts, where the verrucae become low and rounded, calices exactly like Vaughan’s figure, others at the same level have only one series of septa which is often incomplete, and, in places, the calices are almost empty. On the other hand, calices on the verrucae show little variation and are as in the A specimens. Form C : The foliose specimen is only 16 cm. high, so none of Vaughan’s regular calices are present on the smooth part, and the septa are very irregular, with a distinct, even massive columella on one side of a branch, which may be absent altogether on the other. Directive septa rare ; higher up many calices practically empty. Again, calices on the verrucae with 12 delicate septa and distinct, though slender columella, which may be detached, or joined to one directive. Distribution : Widely distributed in the Pacific ; numerous records from Java to Tahaiti, the Marquesas and Hawaii (Vaughan, 1907, under the name P. modumanensis) ; but the only record from the Indian Ocean is from Cocos Keeling Atoll. I have no speci- mens from Mauritius, and there is no record from the Red Sea. Curiously Faustino does not mention it from the Philippines. [Family Astraeidae : species with distinct calices and cyclic septa.] Genus Leptastrea. Stephenson and others (1931) give four references to the genus : swamp passages on roots, etc., the reef flat and mangrove park ; the anchorage ; Batt Reef ; and dredged at Stn. 27. S. Manton records tiny colonies of L. roissyana in the moat, becoming larger in the deeper part, on the descent of the seaward slope (scarce and small). Also on Traverse II. On Traverse III larger specimens, L. ehrenbergana (identified from a boat). On p. 300 is a note on the wide range of conditions the species will tolerate. It occurs also on the crest of Yonge Reef in the form ehrenbergana (as on the barrier edge in Tahaiti), and on Batt Reef. After my experience in Tahaiti and the Red Sea I am unable to divide this genus into the species ehrenbergana, roissyana or transversa, since all grade into, and are all derived, by simplification, from L. purpurea, which is the prior name (Crossland, 1931 and 1935). However, I retain the names purpurea, ehrenbergana transversa and bottae as a matter of convenience, and for reference to the ecological papers. Like Vaughan, I prefer the name transversa Klz. to roissyana M. E. and H., since it is far from certain what the latter is, though Vaughan shows (p. 92) that it is near to the MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 115 form purpurea. Klunzinger’s illustration, fortunately a photographic print, shows an extreme form, but this grades into more ordinary forms not only hi Tahaiti but also here, in spec. D.8.E., in which typical " transverse ” calices occur here and there. Leptastrea purpurea Dana. (Plate I, fig. 5 ; Plate III, fig. 3.) 1918. Vaughan, p. 91, pi. 30, figs. 1, la (showing Dana’s type). This is the form from which the others are derived. A typical example was brought from Low Isles by Mr. G. \V. Otter, and two very small fragments are labelled “ T.l. moat ” ; it is much more rare than its derivatives, and I have not seen it in Tahaiti or in the Red Sea, though there some ehrenbergana specimens approach it. I include here a very curious specimen, Xo. B.M. 420, dredged from Stn. 27, which is unlike any other specimen or figure I have seen, though nearest to Vaughan’s figure of Dana’s type (1918, pi. 30, fig. la). The coral forms a crust, 10 mm. and more thick, on a dead Favia, the living area rounded-triangular, 7 cm. across the base X 7 cm. base to apex. The surface is undulating, but remarkable in this genus for its smoothness. The calices are very shallow, being filled in, except just over and near the columella, by the thick, and for the most part, horizontal septa ; their inner sides are nearly vertical. Consequently to the naked eye the coral resembles a Siderastrea, the characters of Leptastrea appearing only under a lens. All the septa have rough sides and denticulate margins ; they are in 4 cycles, primaries prominent and very thick, only the last cycle thin, so that the calices are almost filled up, the loculi being exceptionally narrow. Columella solid, sometimes oval or round, sometimes compressed, bearing small but distinct tubercles, which may be in a roimd or oval group or in rows, often a single row, according to the shape of the columella. Leptastrea ehrenbergana M. E. and H. Two specimens, one a mere scrap, the other 14 cm. x 10 cm. x 7 cm., showing much less irregularity of shape, and even of calices, than those of the Tahaitian barrier. In the small piece from Batt Reef the roughness of the sides and the granulations of the septa vary remarkably, in some cases being very marked, in others almost smoothed out. Some septa hi the larger specimen from the moat at Low Isles, are clearly “ swollen in the calyx ” ; in the smaller they are “ swollen in the theca,” as Vaughan remarks of Matthai’s pi. 17, fig. 6 (1914). I have a note that this distinction between “ species ” was not made out in my 50 Tahaitian specimens. Leptastrea transversa Klz. (Plate LIV, figs. 1-3.) Specimen 227 is remarkable for its round, separated calices in which cyclical arrange- ment of the septa can hardly be seen, and, in short, the specimen would appear to be an Orbicella or Favia. It grows as a crust over a dead Favia, 0 to 25 mm. thick, and 11 cm. x 13 cm. in area, and almost quite smooth. Multiplication by budding cannot be made out on the main area. However, the following features indicate an abnormal specimen of Leptastrea transversa ; smooth intercalicinal grooves, over which the costae do not meet, smoothness of the upper part of the septa, and the way they run over the tops of the 116 GREAT BARRIER REEF EXPEDITION columellae, and make a bilateral symmetry, as in fig. 2, on Plate LIV. The columella itself bears tubercles, quite characteristic of Leptastrea, and is often compressed, but how disguised are the Leptastrea characters is shown by the photographs on Plate LIV. I take it that the main area of the specimen was in nature a vertical face, probably crowded by other corals, and the edge turned over the block of Favia is the upper horizontal surface, growing in somewhat more normal conditions. Here the smoothness of the side area is less marked, budding is frequent, the walls are thin, so calices polygonal, the intercorallite grooves mere notches in the costae, columellae often non-existent, septa thin and smooth on sides and edges, or granulated on both, with small teeth only near the columella. Leptastrea bottae M. E. and H. (Plate I, fig. 4 ; Plate II, figs. 2, 3.) 1879. L. inaequalis Klunzinger, Taf. v, fig. 6. 1914. L. solida Matthai, p. 69, pi. 18, fig. 5. 1918. L. bottae Vaughan, p. 94, pi. 31, figs. 3 and 4. 1936. L. bottae Yabe, Sugiyama and Eguchi, p. 27, pi. 30, fig. 1. I agree with Vaughan that Matthai preferred the name solida to bottae on insufficient grounds. Yabe, Sugiyama and Eguchi are the only later authors to mention the species ; they give no reference to Matthai or reasons for preferring the name bottae. The present specimen, No. 407, resembles Klunzinger’s (1879a, taf. v) fig. 6, and Matthai’s (1914, pi. 18) fig. 5. In the small crowded cahces of the more usual size, only the six thick primary septa reach the columella, or the secondaries may reach it deep down in the calyx, but generally they are small ; tertiaries are just visible or are absent, but their costae, low and rounded like those of the other series, are generally present. Columella greatly reduced, but may bear vertical points, and septa often bear paliform lobes. As seems to be usual in this species, giant cahces are present ; in these, numerous septa reach the tuberculated columella, which seems to block the bottom of the theca. Comparison with the other species, and with an intermediate specimen in the Kobenhavn museum, indicates that these “ giant ” cahces are, in fact, nearer to the normal form, the more numerous and smaller being the farthest from the ancestral type. A longitudinal section of this species has not yet been figured ; it is remarkable for beams* connecting the thecal walls, some solid, some hollow. I give a photograph on Plate II of a broken, not cut and ground section, broken sections being generally the more instructive, and, indeed, a cut and ground section might not show these curious structures. Compare Milne Edwards and Haimes’ (1848) pi. 9, fig. 3a (for Phymastrea valenciennesii). Leptastrea pruinosa,\ sp. n. (Plate III, fig. 1.) The minute specimen,! B.M. 414 (part), has no locality, but its being on a fragment of Lithothamnia suggests that it was dredged. It consists of 6 full-sized cahces and 6 small, forming a crust 15 mm. x 7 mm. * Visible also at surface. — [A.K.T.] f Latin, pruina, a white frost, referring to the spinules on the septa. j This bolotype now bears B.M. Register No. 1934.5.14.630. It appears possible that it may have been separated from 1934.5.14.414, Acropora squamosa. This came from Lizard Id., A. Reef, and is marked a396. — [A.K.T.] ilADREPORARIA, HYDROCORALLIXAE. HELIOPORA AND TUBIP OR A— CROSSLAND 117 Calices are between round and polygonal (see Plate III) the larger 4 mm. to 5 mm. across. Lines between them quite smooth. Septa, but not walls, exsert to about 0-5 mm. They are decidedly thick oyer the walls, thinner in the calices, where they appear much thicker than they really are through the number of square-ended spinules they bear on their sides. In some cases these meet oyer the top of the septum, forming a little transverse lappet, but in others the thin edge of the septum, finely toothed and bearing- only minute lateral spinules, projects above the part thickened by the massive spinules. These are smaller or absent also over the walls. At the edges of the crust the septa project as costae, smooth or very lightly spinulose. Septa in three cycles, with traces of a fourth, as thin short septa or mere platelets on the walls. The columella is small, almost rudimentary, the tubercles of the upper surface 3 to 6 in number, spinulose hke the innermost teeth of the septa ; this makes another difference from other species. I have long hesitated to describe so small a corallum of a genus in which all the species so grade into each other, but it is very clearly different from any other specimen of any species of the genus. Genus Cyphastrea. The genus is of universal distribution, but it is interesting to find that the ratio of the species is different in different regions. Here we have 10 specimens presumably taken at random, of which 8 are C. chalcidicum and 1 each C. serailia and C. microphthalma. Contrast the northern Red Sea where all these three species are common, but C. microph- thalma decidedly the most abundant. Gardiner found it not abundant in the Maldives, but his collection contains 2 of C. serailia, 4 of C. chalcidicum, and only 3 of C. microph- thalma. It has not been recorded from the Great Barrier before, but it is evidently common, since Stephenson and others give five references, finding it on roots, etc., in passages in the mangrove swamp, scattered small colonies in the mangrove park and Thalamita flat, as well as on the seaward slopes and the anchorage. It is also recorded from Yonge and Batt reefs. S. Manton refers to it on six pages and in two diagrams. There is a good deal of variation in all these specimens, but not the extremes found in Forskaal’s collection. (Crossland, 1941.) Cyphastrea chalcidicum (Forsk.). The 8 specimens are divisible into two sets, without intermediates. A. Humpy, rough with projecting thecae, such as shown by Matthai (1914), pi. 13, fig. 7 (for M. microphthalma ), or pi. 38, fig. 5 (for C. serailia). B. Smooth, thecae hardly projecting, with appearance of lines between the calices, like Matthai’s pi. 14, fig. 1, for this species, or pi. 13, fig. 8 (for C. serailia). Set A are numbers 285, 44, 5, and B.M. 375. The last specimen is from Batt Reef, S.W. side of S.E. corner, wave zone. This specimen is unattached, and growing on all sides, so cannot have been exposed to very heavy waves. Set B are 198, 17, and 43. The last is the largest specimen, 15 cm. x 10 cm. x 7 cm. high, vi, 3. 9 118 GREAT BARRIER REEF EXPEDITION Distribution : Northern Red Sea, through the Indian Ocean to the Philippines. Not recorded from Fiji or Samoa, and does not occur in Tahaiti. C. ocellina, the only species in Hawaii, is very close to this, but Vaughan says it is not identical, so that records of C. chalcidicum from Hawaii must be received with caution. Cyphastrea serailia (Forsk.). A small irregular colony 7 cm. x 5 cm. x 2 cm., encrusting an irregular calcareous mass of uncertain origin. This species and the preceding are often scarcely separable, but this specimen is clearly differentiated by its rather thick spiny septa, with costae of the first two orders not nearly so prominent as in C. chalcidicum, and those of the third nearly equalling them. Inner moat, June reef, No. 415, B.M. 393. Distribution : From the northern Ked Sea, through the Indian Ocean to Murray Island and the Philippines. Not recorded from Samoa. Does not occur in Tahaiti. Cyphastrea microphthalma (Lamk.). A single quite typical specimen, with thick spiny septa, etc. Half the area is dead, the interior extremely rotten, possibly in accordance with the fact that it was dredged from Stn. XXIV. Distribution : From the northern Red Sea to Tahaiti, where it is the only species, and shows reduced growth. Genus Echinopora. I follow Matthai, and others, in placing Echinopora in a division of the Astraeidae, near Cyphastrea (indeed Echinopora is almost a Cyphastrea enlarged) and leave Echino- phyllia in the Fungiidae. Horst (1921, p. 29) says : “ Verrill places the genus Mycedium in the family Echinoporidae and he is right in my opinion.” Umbgrove (1939) follows Yabe, Sugiyama and Eguchi (1936, p. 48) in making a family Echinoporidae to include this genus with Mycedium and Echinophyllia (under the name Oxyphyllia). I do not propose to discuss this arrangement in detail, since the fact that Echinopora has a cyclical arrange- ment of the septa and the other genera have not is against any near relationship. Also, as pointed out, if the leafy Echinopora must join the leafy Fungiidae, then so must also the solid Cyphastrea. Stephenson, Stephenson, Tandy and Spender refer to the genus on pp. 67, 86, 87. They found species at Low Isles, on the seaward slopes and in the anchorage, on vertical or overhanging surfaces below low water. Also on Yonge Reef and Lizard Island, in the latter among other delicate forms on and between larger coral masses ( Porites ) in some depth of water on the edge of the reef. S. Manton (p. 295, pis. xi and xii) found E. lamellosa and another coarser species (? No. 420 below) infrequently in calm water on the seaward slopes of Traverse II, not far from the reef edge, and (pi. xii) on a coral head in water about 9 feet deep. The contrast with the occurrence of E. gemmacea at Ghardaqa in the Red Sea is marked, as this, in the encrusting form, is abundant on the surface of the harbour reefs. Distribution : The genus does not occur in Tahaiti, nor is it recorded by Hoffmeister from Samoa, though Verrill says his E. elegans (E. lamellosa) is from Samoa. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA — OROSSLAND 119 Echinopora lamellosa (Esper). E. elegans and E. concinna Verrill, 1901. Tr. Conn. Acad. Arts. Sci. vol. 11, pp. 138-39, pi. 29, figs. 3, 4. All the 9 pieces of encrusting or free Echinopora have been given this name by Matthai, in spite of very great differences. To the naked eye they fall clearly into two sets : (1) Six pieces ; the surfaces and calices have a soft appearance owing to the number of delicate spines borne in rows on the peritheca, costae, and exsert ends of the septa. These can hardly be shown in a photograph, certainly not in one taken vertically to the surface. (2) Three pieces, of much coarser surface, in which the peritheca bears few and stoutish spines, but is covered with low ridges, while the exsertions of the septa bear coarse teeth. The coarser form comes near to E. hirsutissima, but, as noted above, these species of Echinopora all have members connecting with the others. After working on the genus in the Red Sea I find it impossible to decide whether all the forms there found should be placed in E. gemmacea, as Matthai considers, or "whether, for instance, Ivlunzinger’s E. carduus may not be distinct. Meanwhile I believe that the four species gemmacea, lamellosa, hirsutis- sima and horrida are distinct, though I am not, in every case, able to separate them with confidence. Such scraps as those before me, or the small pieces seen by other authors, are of little help, and a series of large complete colonies is necessary, examined on the reef as well as in the laboratory ashore. Nevertheless the variations among the six more or less typical E. lamellosa are so curious as to deserve mention. No. 125 : Surface nearly flat, very finely “ woolly ” with sp mules. Calices with openings 3 mm. ; if projecting at all only 1 to 2 mm., usually less, and many are completely immersed ; these low walls are thick and rounded ; base larger than the opening. In many of them the larger part of the exsert portion of the septa, within the notch, lies inside the calyx instead of being over the edge, so these appear like huge pali ; the true pah, much smaller, can often be seen within these. No. 27 resembles 125 but is a thin plate much crinkled, and, on the ridges, calices project to 1 mm. No. 456 : Very finely “ woolly.” Calices much larger than in the others, viz., 5 mm. across. Septa all thin, though primaries less so than secondaries, and these than tertiaries ; only slightly exsert, and extending horizontally inwards as far as the edge of the columella, to which they drop vertically 1 mm. or less, hence to the naked eye the calices look like truncated solid cylinders. No. 38 : Lamina thinner than the preceding, spinules thicker and coarser. Calices 3 mm. across, slightly more at bases. Primary septa very thick and rough, secondaries less so, tertiaries thin and nearly smooth. No. 24 : Lamina thin and deeply folded, forming in one part a flattened column which, with a little further flattening and fusion, would make a two-sided plate. Calices widely spaced, projecting about 0-5 mm., diameter generally only 2 mm., and, therefore, the tertiaries small and their cycle incomplete. Sides of septa and top of columella spinulose as usual. Primary septa thicker than secondaries. No. 8 closely resembles No. 24. Of the second set, No. 420 is a thick and heavy crust, the peritheca being quite solid, 20 to 25 mm, thick, with a thin free expansion at one end ; the attached thicker part 120 GREAT BARRIER REEF EXPEDITION bears what may be described as narrow humps or incipient branches, very like those figured by me in 1935 (pi. ii, fig. 3) which is indeed a very similar specimen in shape as well as surface. Apparently the characters of the calices were also similar, as I have a note that they considerably resemble those of E. lamellosa, though typical E. lamellosa, with the characteristic folded margins, does not occur in the Red Sea. Peritheca with low rounded ridges, spines few, low and blunt, but spinulose. Calices comparatively large, 4 mm., round or slightly oval. Thecal walls project at most 0-5 mm. but primary and secondary septa exsert 1 mm. above that. Loculi between septa distinct and deep, thus differing much from preceding set of specimens, since, in spite of the large size of the calices, the secondary septa are thin and the tertiaries rudimentary. Columella well developed, dense, one-third of the diameter of calyx, with upstanding trabecular ends. Pali small and not easily distinguished from the trabecular ends on columella. The calices are the same on the humps as between them and on the flat expansion. Specimen 25 is a fragment of the edge of a thin plate, thinner than the expanded part of No. 420. Neither of these are at all folded. It is very like the preceding, but the primary septa are little, if at all, thicker than the secondary, and both are much less exsert. No. 422 is a solid-walled heavy tube, like part of Dana’s (1848) fig. 4, pi. 17, * at the thicker broken end 20 mm. across, the irregular cavity being 7 mm. x 5 mm. Maximum ex- ternal diameter of tube is about half-way up, 32 mm. At the top (the tube is 80 mm. long) the walls thin out to a sharp edge, enclosing a cavity roughly oval, 15 mm. x 6 mm. Calices, septa and perithecal spines like 420, but the spinulae on the sides of the septa are very fine. There are two similar but smaller bulbous tubes with the first set of specimens, resembling them in, e.g., well-marked ridges on the peritheca. Distribution : Not in the Red Sea, but Indian and Pacific Oceans as far as the Philippines and Fiji. Echinopora horrida Dana. (Plate IV, fig. 3.) Umbgrove, 1939, p. 39, pis. vii and viii. Sample No. 26, No. B.M. 285, and a number of fragments are labelled E. gemmacea by Matthai. I have already recorded the difficulty of dividing the species of this genus, while still believing them to be distinct ; to quote Umbgrove (1939, p. 39), “ if they are it seems to me that the branched type described here is closely allied to E. gemmacea. It may be that it is nothing but a variety or a growth form of that species. Since I have no convincing proof for this opinion I maintain the name E. horrida Dana. . . . The question can be settled only by growth experiments on the living reef. I collected no specimens resembling Echinopora fruticulosa Ehrb. as figured by Klunzinger in his plate 6 fig. 4, and by Crossland ( Proc . Zool. Soc. London, 1935, pi. iii), all the specimens from the Bay of Batavia having branchlets with a laciniate and alate apex.” With this I agree entirely only adding that nothing comparable to this form has been seen among the many speci- mens of E. fruticulosa I have seen in the Red Sea, and that a careful, special search of the reefs of Malaysia or the Great Barrier Reef might settle the relationships of E. horrida to, e.g., E. lamellosa and E. hirsutissima independently of growth experiments ; * Crossland had omitted the date and numbers in his typescript. This is the only figure by Dana of which I know that matches the specimen. — [A.K.T.] MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AXX> TUBIPORA — CROSSLAXD 121 but relationship to E. gernmacea is ruled out by the distinctiveness of the fruticulose form of that species. All the published figures of E. horrida show that while E. gernmacea, and probably the other species have quite distinct fruticulose forms, E. horrida has generally no such distinction, all specimens but this of the Great Barrier Reef being mixed lamellate and branched. The specimen illustrated on Plate IV is quite clearly the same as that on Umbgrove’s pi. vii, the only difference being in the absence of expansions above noted, the proportion- ally few laciniate branches and the great development of heavy branches with conical thecae. Of the former types most have evidently been broken off. It seems likely that these heavy conical calices have a physiological origin, since strikingly similar forms are foimd on the lower branches of Oculina spp. from Bermuda, as shown by Verrill, (1901), pi. xxxii. Nothing similar has been seen in any specimens of E. ge?nmacea, fruticulose or otherwise. I quote Umbgrove again : “ From the facts mentioned here it is evident that there is no reason to follow AVells in his instituting a new generic name for E. horrida.” Distribution : Malaysia, Philippines, Fiji. [Family Oculentdae.] Genus Madrepora. Vaughan, 1907, p. 80. The use of this name for an Oculinid, for which Verrill is responsible, is a regrettable example of legalism versus common sense, but it is too late now to go back to the old well-established nomenclature. Vaughan (p. 82) gives his researches into the meanings of Amphihelia, Diplohelia and so on, and ends : “ However, I believe that pointing out the tremendous confusion of Duncan’s work, and by attacking a name to something definite, a start may be made to unravelling the tangle of the Amphiheliae ( Madreporidae ) striatae.” To go back into this ancient jungle would be merely obstructive, and it is to be hoped that Vaughan’s “ name for something definite ” will not be disturbed, but, under the present legal anarchy, any- thing may happen to any name. Madrepora kauaiensis Vaughan. Vaughan, 1907, p. 81, pi. viii, figs. 1, 2-2a and 3. Three little fragments of thin twigs were dredged from Stn. XV. They correspond exactly with the thin branches of Vaughan’s Hawaiian specimen except that the costal striations of the stems are, in places, less prominent. The species is clearly distinct from Moseley’s two from the East Indian Archipelago and the Philippines. All Moseley’s and Vaughan’s specimens are from deep water ; this is the only record from comparatively shallow water. Distribution : Hawaii, 5 “ Mabahiss ” Stations in Indian Ocean. Genus Galaxea. Stephenson and others (1931) refer to the genus on pp. 44, 67, 86, 87 and 90. One of the corals mentioned as characteristic of the reef flat, of the seaward slopes 122 GREAT BARRIER REEE EXPEDITION and anchorage ; conspicuous on Yonge Reef, the fringing reefs of Lizard Island and on Batt Reef. S. Manton refers to the genus on pp. 289, 295, 300, 302, 303, and in pis. xi, xiii, xvi. Species occur on deeper parts of the moat and are dominant on the deeper parts of seaward slopes on Traverses I and II. Here specimens attain a size of over 10 inches. Compare the Red Sea where cushions 4 to 6 in. across are not separate colonies but the tops of columns forming very large masses below. Apparently such masses do not occur in the Great Barrier region. The colonies shown on the plates are minute. \Galaxea fascicularis (Linn.)]. The four specimens, only one of which [G. musicalis ] is more than a scrap, have been labelled G. fascicularis (Linn.) and G. musicalis (Linn.) by Matthai. [Galaxea clavus Dana], On this latter name Vaughan (1918, p. 100) remarks : “ Regarding the name musicalis which Milne Edwards and Haine applied to the species and attribute with a query to Linnaeus, it is not known, and probably never will be known, what species Linnaeus meant. Because of this uncertainty the name clavus proposed by Dana, concerning which there is no doubt, should be applied.” My recent work on Forskaal’s corals, and other species of the genus in Kobenhavn Museum, make me share to the full Vaughan’s doubt as to the propriety of following either Linnaeus or Milne Edwards and Haime, and I therefore regard clavus as the first real name for this species. In my edition of Forskaal’s work I have shown how certain characters usually taken as diagnostic are only characteristic of stages of growth, and especially of dwarf specimens from the reef flat. Such are a continuous peritheca with comparatively thick-walled vesicles, the divergence of calices and their narrowing to their bases. These two last can only be found near the base of a new column, or when new buds arise, and the amount of the projection of the calices may depend on the rate of growth of the peritheca at the time. The number of septal cycles naturally varies with the size and development of the calyx ; there may be five in G. fascicularis, but often only three, with or without an incomplete fourth. The thickening of the septa in the walls is not a constant difference, though well marked in some specimens of G. clavus, e.g., Matthai (1914, pi. xvi, fig. 2). The distinctions between the species are (1) their growth forms, (2) the sizes of the calices, (3) the columellae. The growth forms are cushions in G. fascicularis, but, except in very shallow water, these cushions are the tops of contiguous columns, which therefore never bear polyps on their sides, whereas in G. clavus the columns are free, and bear polyps on all sides. This difference is quite real, since free columns have never been seen in the Red Sea, where I have seen many hundreds of specimens of G. fascicularis, both living and semi-fossil. It is in the latter, of course, that the internal arrangement of big colonies is easily seen to be an agglomeration of columns. The sizes of the calices grade into one another, as would be expected, e.g., specimen 188 T 1, labelled G. musicalis by Matthai, has occasional calices as large as the average of No. 40, G. fascicularis. The third distinction is given by Matthai as “ columella poorly developed ” and “ columella distinct ” for the two species. Milne Edwards and Haime say “ columelle nulle ” for the synonymous G. irregularis. Defining a columella as a structure formed by outgrowths from the inner VADREPORARIA, HYDR OCOR ALLENAE , HELIOPORA AND TUBIPORA— CROSSLAND 123 margins of the septa, then, though the septa meet centrally, there is, in G. fascicularis, as a rule no columella at all, at most a mere rudiment through which the septa pass, whereas in G. davus, though small, it is distinct. Vaughan (1918), p. 99, remarks: "'Although large colonies [of G. clams ] usually form ascending columns, they do not invariably do so. There is in the U.S. National Museum a colony 15-5 x 18-5 cm. in diameter at the base, which has a nodulate upper surface, but there are no columns.'’ Naturally columns will not be formed in shallow water, and it is from reef pools that so much of our material conies. Specimen 188 of this collection is incomplete, but measures 14 cm. x 9 cm. X 5 cm. thick, is almost quite flat, sloping down round the sides ; it is from shallow water. S. Manton (p. 288), writing on Traverse I, finds G. musicalis, 3 to 7 inches across, which becomes large and abimdant beyond 886 feet, and after 945 feet the large-polyped G. fascicularis is usual (see graph 29), but apparently all specimens are really extremely small. Distribution : G. fascicularis, the only species in the Red Sea, so far as is yet known, extends through the Indian Ocean and Pacific as far out as Samoa. G. clams, Indian Ocean and Pacific as far east as Fiji, but not in Samoa, nor Tahaiti, nor in the Red Sea. Acrohelia horrescens (Dana). Acrhelia horrescens Yonge, 1930, vol. i, No. 2, p. 17. This is one of the species which are common in certain places, absent or rare in others of the same district. It is not in the collections from Low Isles or the vicinity, but both Mayor and Yonge found it in the Murray Islands, and Yonge in the Capricorns ; it is common in Iwayama Bay, Palao Islands. Umbgrove (1939) also writes (p. 12) : “ Thus, e.y., not a single specimen of Acrohelia horrescens was found hi the Bay of Batavia although this species abounds on the Togian reefs.” On p. 14 he gives a long list of corals from Amboina (not only Bedot’s collection) from which A. horrescens is absent. It is close to the genus Galaxea, and, judging from certain specimens of a second species of the genus dredged by Dr. Mortensen off Banda, the two genera may be really identical. This view is supported by Yonge’s remark, “ The structure and behaviour of the polyp is identical with that of Galaxea and description is therefore unnecessary, particularly in view of the possibility that this genus should properly be placed in the Orbicellidae.” I adhere to Matthai’s division of the Astraeidae, which includes Orbicella, though I do not follow him in merging the genus with Favia, since, as I have shown, there is the closest relationship between Orbicella and Favia. Acrohelia should join the Astraeidae beside Galaxea. Distribution : Malay region and as far east as Fiji. [Sub-family Montastreinae.] Genus Orbicella. I have no hesitation in following Gardiner and Vaughan, and in fact all authors but Matthai, in keeping this genus distinct from Favia. Not only are the round calices accompanied by intercalicinal budding sufficient grounds for doing so, but Matthai is hasty in inferring that cyclical arrangement of the septa is absent in the species which he removes to Favia. As Vaughan notes in 1914 and C. Crossland in 1931, cyclical arrange- 124 GREAT BARRIER REEF EXPEDITION ment is often to be made out, and, in some specimens, is the rule. It is clear in some calices of Klunzinger’s 0. laxa (one of the species removed to Favia by Matthai), but this is only one example, and I particularly refer to my F. ingolfi of Tahaiti, and the wish expressed in 1931 and 1935 that its polyps, which are preserved in the British Museum, might be sectioned. Other examples are given under 0. vacua sp. n., below. I find the four Indo-Pacific species difficult to define, and I am uncertain whether Matthai’s F. laxa is really the same as Klunzinger’s 0. laxa (see note on p.125 ). In Tahaiti the three species, versipora, wakayana ( i.e ., curta) and solida definitely do merge, but, as I showed in 1935 for other species, this does not necessarily apply to other regions than Tahaiti. Gardiner (1904(6), p. 774) writes, “ I must still assert my inability to find characters which clearly separate Plesiastraea and Leptastraea from this genus (P.Z.S. 1899, p. 751).” A case of a very striking resemblance of a spechnen of Leptastrea to Orbicella versipora has been given on p. 31. Note also the strong probability that 0. mammillosa Klz. is a form of Echinopora gemmacea. At the same time the facts that cyclical arrangement of septa is not universal, and that budding may be accompanied by fission, show the close relationship to Favia, and that the two divisions of the Astraeidae as tabulated by Matthai are not hard and fast. The following are the species of this genus in the Indo-Pacific area : 0. curta, 0. ingolfi, and 0. vacua n. sp. ; while the following may all be synonymous with 0. curta, viz., 0. versipora, 0. laxa and 0. solida. Orbicella curta Dana. 1899. Gardiner, Orbicella wakayana, p. 753, pi. 49, fig. 2. 1914. Matthai, G., Favia wakayana, p. 104, pi. 25, fig. 4. 1918. Vaughan, T. W., Orbicella curta, p. 86, pi. 28, figs. 2 to 5. 1931. Crossland, C., Favia versipora, p. 384, pis. 15 to 19. Matthai, examining Gardiner’s collections in 1914, decides that four of his species of Orbicella are the same as his Favia wakayana. He remarks that “ it is likely that the present species may have been previously recorded by Dana, but this point cannot be settled till Dana’s Astraeid types are examined.” This was done by Vaughan in 1918, who shows that this species is Dana’s 0. curta and 0. coronata, the former name taking precedence, and publishes photographs of both types. Specimen 401 is labelled “ ? Favia wakayana ” by Matthai but I see no need for the ? mark. The scrap labelled “ Orbicella A, Nigger Head, Undine Reef,” but not numbered, is the same species. From S. Manton’s reference, p. 305, pis. xiv and xv, the species is rare on Low Isles, more common on the outer reefs. Distribution : If distinct from 0. versipora, Pacific Ocean only, but as far east as the Tuamotu Atolls, i.e., the limit of the Indo-Pacific area. Orbicella vacua sp. n. (Plate II, fig. 1, 4 ; Plate III, fig. 2.) Three small specimens I am unable to place in any known species. They are sample numbers 399 (5)* ; 400, “ Favia 2. Orbicella B,” from reef crest, June Reef ; and 433, * This probably means “ Favia 5.” — [A.K.T.] MADRE POR ARIA. HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 125 "outer barrier, June Reef, digger Head. " The first two are small crusts, the third a portion of a solid cushion at least 75 mm. thick. This species resembles 0. curta strongly at first sight, having similar exsert septa and costae bearing the rather characteristic transverse lappets. Usually the walls project more, and the costae are more prominent, but in much of the area of No. 399 they are no more so than in most specimens of 0. curta. All the septa are narrow, and descend vertically into a deep calyx, and there are no traces of pah — hence the specific name. The Tahaitian specimens of 0. curta usually lack pah (Crossland, 1931, p, 384, pis. 15 to 19), only traces being found in a few calices of 8 specimens out of 60, but this is the result of degeneration, as shown by the long teeth of the lower parts of the septa and the loosely made columeha. In 0. vacua the upper parts of the septa are merely spinulose, the lower either entire or with short blunt teeth, which abruptly merge into the sohdly made columella, on which only the points of the trabeculae are usually visible, the heavy beams being often fused together. No. 433 is labelled ? Favia laxa (Klimzinger) by Matthai but the specimen differs from this species as much as it does from 0. curta. In fact Orbicella laxa Klz. and 0. curta are almost certainly the same. I presume that Klunzinger’s and Matthai’ s species are the same, though Klunzinger (1879a, p. 50) says “ Knospung fast immer extra - calicinal,” while Matthai (p. 100) says “multiplication by equal or subequal fission.” I also find several calices in Klunzinger’s photograph showing distinct cycles of septa, which are not to be seen hi Matthai’s, but this is not even a specific difference in either O. curta or O. vacua. Klunzinger describes the calices as 8-10 mm. in diameter, but they average 6 mm. with a maximum of 8 mm. in the photograph, as in Matthai’s. They may be oval in Klunzinger’s species, but not as often, nor so greatly, compressed as in Matthai’s. Marenzeller gave the name 0. laxa to several large specimens from the Red Sea, but neither he nor Matthai describe them. The contrast with 0. curta, normal or Tahaitian, is complete. Multiplication, in these three specimens, is by extracalicinal budding alone, and all calices are round except a few in depressions in 399. In No. 433 four cycles of septa are seen in a number of calices with almost diagram- matic clearness ; in 399 they are less evident, all the larger septa are alike and all reach the columella, except occasionally one joins another. The quaternaries are very small, sometimes only their costae present, alternating with the others as in Favia. Owing to these differences these three specimens together must be regarded as the syntypes* of the species. They must not be separated. [Family Astraeidae : Species Without Cyclical Septa.] Favia favus (Forsk). 1901. F. affinis Marenzeller, p. 119. 1906. F. savignyi Marenzeller, p. 82, pi. 25. 1914. Facia favus Matthai, p. 79 (with synonymy), pis. 21, 22, 36. Six specimens are definitely given this name by Matthai, viz., samples 3, 134, 149, * Crossland wrote “ type.” But in accordance with the recommendations on International Rules of Nomenclature, B.M. Register No. 129, is hereby designated holotype, and Nos. 370, 396 paratypes : Crossland (1941, p. 10) remarked that such a proceeding, i.e., designation of “ type and cotypes ” (as he called them), though scientifically absurd, had its practical use. — [A.K.T.] 126 GREAT BARRIER REEF EXPEDITION 160, 169, 170, and two more marked with a query. Of these latter I find No. 150 indis- tinguishable except that the number of septa continuous over the walls is greater than usual, and No. 410 I propose to treat as a distinct variety, var. crassidens. S. Manton refers to it on pp. 285, 286, 288, 294, 298, and 306, and on plates viii, ix, xi, xii and xvi, from which I conclude that the species is one of the commonest of the Faviidae on the Great Barrier Reef as it is throughout the Red Sea. On Traverse I, were numerous small colonies, 2 to 7 inches across ; in deeper parts, between 888 and 1014 feet, the commonest species occurring at the rate of 10 per square yard. In all these shallows the specimens are small, in the deeper water, according to the plates, it is rare and smaller than most other Faviidae ; on the inner part of Yonge Reef one irregular colony of 27 cm. is shown. In the Red Sea I do not remember any very large specimens, and the abundance of speci- mens in museums I believe to be due to this habit of forming small, nicely rounded growths in shallows. Onp. 306 S. Manton writes, “ Fungias and F. favus are more particular in habitat, and occur here and on coral heads and leeward slopes and moats of Low Isles, but are not frequent on other parts of Yonge Reef.” The specimens are matched by published figures as follows : No. 149 : Matthai, G., 1914, pi. 20, fig. 6, “ Red Sea, typical form ” ; pi. 22, fig. 3, Forskaal’s type, typical : and fig. 5, his type of M. cavernosa. No. 150 : As above, but there are frequent meetings of septa over walls. No. 170 : A mere scrap, like above but distorted. No. 160 : As No. 150. No. 169 is intermediate between typical forms and that shown in Matthai, 1914, pi. 21, fig. 1. No. 134 : Marenzeller (1906), pi. 25, fig. 86. No. 3 : Matthai (1914), pi. 21, fig. 1 ; and Marenzeller (1906) pi. 25, fig. 84. Distribution : Common in the Red Sea, including the northern part ; recorded from all over the Indian Ocean. Pacific records are few, Tongatabu ; Philippines and Samoa ; and now the Great Barrier Reef, but not from Murray Island. It does not occur in Tahaiti, but a specimen from the Tuamotu Atolls in the Museum at Papeete is probably of this species. F avia favus var. crassidens var. n. (Plate XIV, fig. 5.) Specimen No. 410, though heavy for its size, is remarkable for the thickness of its septa rather than of its walls, in fact the calices are near together in comparison with Matthai’s (1914) var. 2, figured on pi. 20, fig. 4, or pi. 21, fig. 3 ; or by Marenzeller (1906), pi. 25, fig. 88. Prominence of a few septa is common in this species, and is shown in Matthai (1914), pi. 21, fig. 3, pi. 22, fig. 1, and, in some degree in Nos. 149 and 134 of the present collection ; but the state of these prominent thickened and exsert septa, as shown on pi. xiv, fig. 5, is unique. To this photograph I refer for numbers and measurements. It is regretted that the specimen should be only a small part of a colony. Favia valenciennesi M. E. and H. 1914. Favia bertholleti (Val.) Matthai, p. 94, pi. 22, fig. 7 ; pi. 23, figs. 4, 6 ; pi. 24, fig. 1. 1918. Favia valenciennesi Vaughan, note 2, p. 100. 1924. F. valenciennesi Matthai, G., p. 14, pi. 4, fig. 1 ; pi. 11, fig. 2. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA — CROSSLAND 127 Specimen 163 is given the name bertholleti by Matthai, with a ? mark on his pencilled label. The species is difficult to define, as shown by the number of photographs given above, and by the fact that Matthai (1914) shows on pi. 22, fig. 7, one of Forskaal’s types of F.favus, " ? F. bertholleti (Yal.) . . . perhaps only a thin-walled F.favus.” S. Manton does not mention the species ; it is evidently rare on the Great Barrier. This specimen shows the same change from moderately deep calices on one side of the colony to very shallow ones on the other that was seen in a specimen of F. doreyensis , but in this case the whole colony was alive and apparently healthy when collected. This difference is common between the upper and lower parts of many corals, though not often so marked. Intercorallite grooves are not present, this being var. 1, the fight form. It is very like Matthai's (1914) pi. 23, fig 6, but the calices are usually rounder. Distribution : Red Sea and Indian Ocean. The present record from the Great Barrier Reef is the first from the Pacific. Favia doreyensis M. E. and H. 1911. Matthai. G., F. doreyensis, M. E. and H., p. 81, pis. 9, 22, 32. 1918. Vaughan, T., W. F. pallida Dana, p. 105, pi. 38, figs. 1-7. 1921. Matthai, G., F. doreyensis, p. 11. Matthai’s Teason for not adopting Dana’s name in 1924 is apparently that Yauglian includes F. hululensis Gard. in F. pallida, which species Matthai considers should be kept distinct. The species hululensis is not present here. The species is peculiar in having round separated calices which divide very unequally and yet produce smooth colonies. One specimen, No. 154, calls for remark ; it is an oval mass, 12 cm. x 8-5 cm. x 6 cm. high, somewhat hollow underneath. About half of it is dead ; along the growing inturned edge it has the characters of F. doreyensis, but, as we pass over the upper surface (probably the vertical surface in life) calices become more and more shallow, their edges less and less distinct, until, over the dead half of the colony, the thecae are quite superficial and widely separated, the result being almost exactly like a figure of F. favosa given by Matthai (1924), pi. 2, fig. 8. Doubtless the same causes, position in the colony and mud or crowding, has had the same effect on these two so different corals, and on the specimen of F. valenciensii referred to above. S. Manton refers to this species on pp. 285, 288, 294 and 297, and on pis. viii, ix and xiii. It is of wide local distribution, but infrequent, and, like all the Low Isles Faviidae, very small. Distribution : Indian Ocean, Pacific to Philippines, Fiji, and Samoa, but not from the Red Sea or Gulf of Aden. Favia speciosa Dana. 1914. Favia clouei (Val.) Matthai, p. 89 ; pi. 10, fig. 6 ; pi. 23, figs. 1, 2, 5 ; pi. 34, fig. 1. 1918. Favia speciosa (Dana) Vaughan, p. 103, pi. 36, figs. 1, 2, 2 a, 3, 4, 4a ; pi. 37, figs. 1 to 4a. 1924. Favia speciosa (Dana) Matthai, p. 12, pi. 1, figs. 3, 5, 8* ; pi. 11, fig. 3. Matthai adopts Vaughan’s correction of the name clouei to speciosa in 1924, but labels the four present specimens as F . clouei. * As this plate is printed upside down figs. 3, 5, 8 of the ‘ explanation of plate ’ appear as figs. 7, 5, 2 on the plate itself. 128 GREAT BARRIER REEF EXPEDITION To F. clouei S. Manton refers on pp. 285, 288, 298, 302 and 305 ; and on pis. iii, viii, xii and xiv. On Traverse I there was one in the moat and a few on the deeper part of the seaward slope. On Traverse III (to windward), on the third 100 feet of the slope, “the deeper water Favias of Traverse 1 continue or appear for the first time,” F. clouei being one now to appear. On p. 302 Manton says, “ The deeper water F. vasta, \i.e., F. virens ] F. clouei, [i.e., F. speciosa] Astreopora~\ and Loboplnyllia have larger polyps than the species found in shallow water ” and withstand mud better. This species has been thoroughly illustrated, as shown by the references above, to which many others may be added including Yabe, Sugiyama and Eguchi. No. 159 is very like Dana’s type of Astrea pandanus figured by Vaughan, T. W. (1918), Pis. 36 and 37, and some of its calices low down on one side resemble both fig. 4 a on pi. 36 and the whole of the flat specimen dredged at Stn. XXIV. This leads on to No. 215, a distorted scrap from “Detailed survey 1. Favia from scanty corals, weed zone, deep water.” Nos. B.M. 419, 418 and 609 dredged at Stn. XXIV, with perfectly flat calices, marked off by thin polygonal lines, may be this species, but certainty is not possible. Distribution : Northern Red Sea through the Indian and Pacific Oceans as far as Fiji and Fanning Islands, but not found in Samoa or Tahaiti. Favia stelligera Dana. F. acropora Matthai, 1914, p. 102. One portion labelled “ No. 49, F. acropora ” by Matthai, represents this species. The names acropora and lobata have the advantage of being descriptive ; in Matthai’s list of 14 references the former is used six times, the latter five times, but Dana is not referred to. In 1918 Vaughan states that the earlier authors’ descriptions are not identifi- able, and have led to error, so that this name is not available for any coral. He therefore substitutes Dana’s name stelligera, remarking that Klunzinger’s pi. 3, fig. 9, of F. lobata might almost have been made from Dana’s type. S. Manton refers once, on p. 289, to the occurrence of this species on the seaward slope of Traverse 1 . As the species is so conspicuous in life, the columnar growths, which are thicker above, and the small calices making it one of the few Astreans certainly recognizable at sight, I conclude that it is rare in the Great Barrier Region. It is not recorded from Murray Island. Distribution : From the northern Red Sea, where it may form large masses, to Tahaiti. Favites halicora (Ehrb.). 1914. Favia halicora (Ehrb.) Matthai, p. 106, pi. 26, figs. 3, 5-7. 1918. Favites halicora (Ehrb.) Vaughan, p. 110, pi. 41, figs. 1, 2 and 3. 1924. Favia halicora (Ehrb.) Matthai, G., p. 17, pi. 1, figs. 4 and 6.* S. Manton refers to this species on pp. 285, 288, 294, 298 and 306 ; and on pis. viii, ix, xv and xvi. It is generally distributed, as very small colonies, on shallows both at Low Isles and Yonge Reef. In the latter locality it grows larger, but nothing of any size is shown on the plates. f Not true of Astreopora. — [C.C.] * As this plate has been printed upside down figs. 4 and 6 of “ Explanation of Plate I ” are really figs. 6 and 4 on the plate. MADREPORARLA HYDEOCORALLIXAE, HELIOPORA AND TTBIPORA— CROSSLAND 129 Four small specimens and a scrap are labelled “ Fa via halicora ” by Matthai. The last, 425, is too small for certain identification. 173 is thin walled, 162 and 180 are decidedly thick walled, and 229 is intermediate. Grooving between the calices is strongly marked in 162 but hardly at all in 180, though both are thick walled. For distinction from allied species, e.g., F. abdita, see Vaughan (1918), pp. 101, 110. For the relationship between thin and thick walls compare Matthai’s two figures in 1924, which are taken from the same specimen, and illustrate the variation with position mentioned above in the cases of e.g., F. doreyensis, F. virens and F. valenciennesi. Distribution : From the northern Red Sea through the Indian Ocean, including the Xatal coast : the Pacific, including Murray Islands, Fanning and Samoa, but not Tahaiti. Favites abdita Ell. and Sol. Of this generally common species only two small pieces are present, but S. Manton refers to it on six pages and six plates. The great likeness of some specimens of F.flexuosa to some of F. abdita is commented upon on p. 46, and by Yabe, Sugiyama and Eguchi, while Vaughan, 1918, p. Ill, remarks : “ Some specimens of F . halicora have a most perplexing resemblance to some specimens of F. abdita (compare pi. 40, fig 4, with plate 41. fig. 2), as Matthai has pointed out. Usually the prominent septal dentitions just within the calices of the former are a good discriminating character.” S. Manton refers to the species on pp. 288, 294, 298, 305 and 307, and on pis. viii, ix, xii xiv and xvi. The specimens noted are quite minute, 1 to 7 inches across, except one shown on pi. xii. This was growing on a pinnacle of dead coral which may be described as small, being 36 cm. across. It is common in shallow water on the two leeward traverses, but not to windward. On Yonge Reef it is found frequently on the iimer parts and occurs also on the reef crest, where the colonies are very small. It is thus one of the species “ able to exist under very variable conditions,” but not able to grow to full size anywhere in the areas seen. Compare this observation with that of Gardiner (1899, p. 758). In Rotuma (?) “ these three species ” (all are F. abdita, fide Matthai) “ live on the extreme breaking edge of the reef, and are exposed at spring tides for two or three hours to the sun. though constantly wetted by the spray. They form also large spreading masses as deep as can be seen outside the reef.” In 1905, p. 787, of P . fusco-viridis Gardiner says, “ Very common on lagoon shoals and outer slope, often forming immense masses both at Minikoi and in the Maldives.” In the Red Sea large specimens occur off a reef edge about Lat. 21° N., but at Ghardaqa in the North it, with F. halicora, is one of the corals which are found as small scattered colonies on the outer part of the shore reef-flat. Large colonies have not been found there. The two specimens are of the blunt-walled variety. Among the numerous illustra- tions published they are matched by Vaughan’s (1918) pi. 40, fig. 3, which shows Dana’s type of A. fusco-viridis, and fig. 5, one of the specimens from Murray Island ; and by Matthai’s (1914) pi. 29, figs. 3 and 4. For illustrations of the more typical growth form see Bedot (1907), pi. 30, fig. 150 (under the name P. robusta), and Matthai (1914), pi. 35, fig. 2. Matthai (1924, pi. iv, fig. 2) shows a curiously branched form from somewhere in the Indian Ocean, which is matched by two in the Kobenhavn Museum from Singapore, in which hillocks become vertical branches. 130 GREAT BARRIER REEF EXPEDITION Distribution : The whole Indo-Pacific, including the Bed Sea, Mauritius and Natal ; but not east of the Fiji Islands. Favites virens Dana. (Plate VI, figs. 1, 2.) 1914. Favia vasta (Klz.) Matthai, G., p. 108, pi. 27, figs. 3, 5, 6. 1918. Favites virens (Dana) Yauglian, p. Ill, pi. 41, figs. 4, 5. 1924. Favia vasta (Klz.) Mattliai, p. 18, pi. 1, fig. 7* ; PI. 11, fig. 1. 1936. Favites fiexuosa (Dana) Yabe, Sugiyama and Eguclii, p. 32, pi. 20, fig. 1. S. Manton refers to this species on five pages and on three plates, using Matthai’s name, Favia vasta. In 1924 Matthai lists Vaughan’s name virens as a synonym, but gives no reason against its adoption, though Vaughan’s reasons appear conclusive ; they depend upon Matthai’s identification of Klunzinger’s Goniastrea halicora with his P. vasta, with which, after deliberation, I agree. I therefore follow Vaughan in adopting Dana’s name. Favites Jlexuosa Dana is raised to a distinct species by Yabe, Sugiyama and Eguchi, though it is given as a synonym of F. abdita by Vaughan. While agreeing with the Japanese authors in its distinctness from abdita, the seven specimens before me bring jlexuosa clearly into the range of F. virens. The distinctions from F. abdita are, however, much less than the resemblances. Yabe, Sugiyama and Eguchi give (1) broader, stouter, and more numerous septa ; (2) broader walls, rounded above ; (3) larger calices. (1) is merely a consequence of the much larger calices, and, in fact, F. virens is only a large form of F. abdita, except that in the former, when septa meet over walls generally the short meet the short, while as a rule the short meet the long in F. abdita. The difference between short and long septa is much more marked in F. virens, the former being decidedly thinner, and extending only a short way down the wall. These small differences suffice to divide the species unless intermediates are found. Just as in F. abdita there are two very well marked forms (1) with thin walls and calices as almost regular pentagons ; (2) with thicker, rounded walls and calices, therefore rounded pentagons or almost oval. The lattef are the F. jlexuosa of Dana and of Yabe, Sugiyama and Eguchi. I regard the columella as a distinctive feature, which is shared with F. abdita, and illustrate it on Plate VI, fig. 1. It is compact and clearly defined ; the trabeculae are delicate and expand into lappets at their ends, the edges of which are spinulose ; these have no resemblance to septal teeth, with which their connection is found only after careful examination. The septa round the columella end in a vertical drop, forming a little pit floored by the top of the columella. Klunzinger’s description as “ grob trabekular ” is thus misleading, as apparently the coarseness refers to the little lappets, not to the columella itself. A lens shows at once that the “ Goniastrea halicora ” figured contradicts the definition of the genus in having a well-developed compact columella. I much doubt whether his var. super jcialis, with its rudimentary columella, formed of ordinary septal reeth, can be the same species. There is nothing like it in the series before me. Each long septum is divided into two parts, the upper narrow, with fine teeth ( i.e ., * As this plate has been printed upside down fig. 7 of “ Explanation of Plate I ” is really fig. 3 on the plate, MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 131 in proportion to the size of the calyx), the lower broader and thicker, with coarser teeth. This lower part, to the naked eve, simulates a crown of pah. The series before me could not have been better selected to illustrate the variation from an abdita-hke form to the very different flexuosa. The points in which the specimens differ are (1) size, (2) depth, (3) walls, thin and sharp or rounded, (4) presence or absence, and depth of grooves between the calices. (1) Diameters of calices : Specimens 174 and 167 have small shallow calices, like those of abdita, but with some larger than are to be found in the latter, viz., 15 mm. in longer diameter. The average and maximum sizes in the other specimens are 15 mm. and 20 mm., but there are only two calices of 20 mm. on one side of the colony. In No. 114 the average and maximum sizes are 20 nun. and 23 mm. In another specimen,* 20 mm. and (one) 22 mm.. In No. AV.2, 20 mm. and 24 mm. (this is the typical flexuosa, almost exactly like Yabe, Sugiyama and Eguchi’s fig. 1, pi. 20), and lastly a specimen* with all calices about 15 mm. in diameter. (2) Depths of calices : in the first two specimens the calices are shallow, especially in the thick-walled No. 167, but in both as usual in all corals, the calices are much deeper on the summits of the dome-shaped colonies. In the third, No. 115, the calices are shallow at one end, 7 mm. deep at the other. No. 114 is similar, but has far larger calices (6 and 7 mm.). In No. 152, all calices are equally deep. In AV.2 all are deep, and in 145 the calices are smaller but about 10 mm. deep. (3) AValls thin and sharp in only the first of the series, and on the flatter part of the fourth. This specimen is particularly interesting, since fairly sharp-walled calices occupy most of the surface, thick walls round the deeper cahces which are marked off by grooves. The fifth and seventh of the series have all the calices separated by grooves, in the latter quite deeply, but in the sixth, No. AA”.2, there are few traces of grooves, i.e., it is a perfect example of F. flexuosa. Illustrations of specimens matching those of this series are provided as follows : (1) No. 174, Klunzinger (1879«), pi. iv, fig. 2 ( Goniastrea halicora). (4) No. 114, the part without grooves between the calices, Vaughan, T. AV. (1918), pi. 41, fig. 5 ; Faustino (1937), pi. 28, fig. 2 ; Mayor (1918), pi. 16, fig. 28. (6) No. AAr.2 (F . flexuosa), ATaughan, T. AV. (1918), pi. 40, fig. 2 ; Yabe, Sugiyama and Eguchi (1936), pi. 20, fig. 1. Of the remainder of this series Nos. 1 14, 115 (the upper part) are illustrated on Plate VI, figs. 1 to 2. The first is remarkable in that the cahces are marked off partly by notches in the septa over the walls, in the usual way, and partly by laminae between the septa parallel to the walls. It is to be remarked that the series would have been even more instructive had larger and more complete colonies been brought home, or, from really large masses several samples been taken with notes on their positions. At the same time I note from S. Manton’s account that, although the colonies belong to a large number of species the colonies are all small — very small compared with those I know in the Red Sea, and those described by * There are seven specimens : Nos. 114, 115, 145, 152, 167, 174, W.2. Crossland did not mention the number of this specimen in his typescript. — [A.K.T.] 132 GREAT BARRIER REEF EXPEDITION Gardiner in the Indian Ocean and Pacific. There are few coral species in which the specific features are constant over any large area of a colony, and lateral crusts are generally abnormal. Distribution : From the northern Bed Sea (where it attains a considerable size) to the Pacific, but not further east than Samoa. Favites aspera Verrill. (Plate V, figs. 1, 2.) 1868. Goniastrea aspera Verrill, p. 32. 1904(6). j F avia par vimurata Gardiner, p. 771, pi. 62, fig. 25. 1914. ? Favia parvimurata Gard., Matthai, p. 113. ? 1918. Acanthastrea hemprichii Vaughan, p. 101. 1927. Goniastrea aspera Verrill, Faustino, p. 141, pi. 33, figs. 1 and 2. 1936. Goniastrea aspera Yabe, Sugiyama and Eguchi, p. 35, pi. xxiv, fig. 3. 1938. Goniastrea aspera Eguchi, p. 350. 1939. Favites aspera (Verrill) Umbgrove, p. 30, pi. iii, fig. 3 ; pi. v, fig. 1. Two specimens of widely different appearance are present, No. 155 being labelled Favia parvimurata, the other No. 166, ? Favia bertholleti, both by Matthai. This latter specimen fits no description or figure of that species, but does connect the heavier No. 155 with the lightly built F. aspera, from which it differs only in having smaller and simpler upwardly directed teeth on the exsert parts of the septa. Gardiner’s original description fits specimen No. 155 better than does Matthai’s, which is abbreviated and omits the important fact that the walls are not only peculiarly delicate but sometimes incomplete, and dismisses the lower septal teeth, carefully described by Gardiner, mere as a conspicuous pali crown. These two points are important also in connection with the much lighter form of No. 166, which is figured with 155 on Plate V, the differences from Gardiner’s specimen being all due to general lightness of build, e.g., the relative inconspicuousness of the lower teeth of the septa and of the toothing of their upper parts. As might be expected, the walls in this example are quite often incomplete, though never wholly missing, and without interrupting the sequence of the septa. The columella is loose, though essentially as in No. 155. For septal numbers, measurements, etc., I refer to the photograph. A complete description of the species can be had by combining the accounts of Gardiner, Faustino (who copies that of Verrill, with notes by Vaughan) and Umbgrove. Vaughan (1918), p. 114, says : “ I doubt G. aspera really being a Goniastrea. Besides having roughly and irregularly dentate septa, the intercorallite walls are often slit, making a combination of characters suggesting affinity with the Coeloria group of Meandra.” Distribution : Maldives, Malaysia, Japan (Yabe, Sugiyama and Eguchi), Southern Philippines and the Great Barrier Beef. Always rare. Genus Goniastrea. The genus is abundant at Low Isles and elsewhere on the Great Barrier, and apparently in many places in the Indo-Pacific it is the commonest of all the Faviidae. The smooth rounded shape and nearly white colour of the commoner species make them conspicuous. Stephenson and others include it in Favia, but S. Manton gives four page references to the genus on four pages and to the several species on numerous pages and plates. At madreporaria, hydrocorallixae, heliopora and tubipora— crossland 133 Hope Reef Hedley and Taylor found it, with Forties, dominating the edge of the reef. This genus has evolved a completely meandrine form, G. benhami Vaughan, noted on p. 136 and figured on Plate VIII, which offers a curious ecological contrast to what we may call the meandrine form of Favia, i.e., Coeloria, in that, while the latter is a more successful species, both in number of specimens and especially in size of colonies than the original form with simple calices, the former is a rarity and always small. Similarly in Lobo- phyllia it is the simple form which is so Important in the Red Sea and elsewhere, the meandrine forms being unimportant. Many species have been attributed to Goniastrea which are so ill described that even their generic position is uncertain. Some of these are tabulated by Matthai (1914, p. 116). It does not seem to me possible, however, to make G. seychellensis Klz. identical with ? F. favus (Forsk.). Quelch’s species are not certainly identifiable, but his G. incrustans has been redescribed by Matthai (1924), p. 21, pi. ii, fig. 4 ; pi. xi, fig. 4. It is very close to G. pectinata. G. lacera Verrill is a Forties* I therefore list only the following 6 species as certainly known : 1. G. retiformis (G. solida or G. parvistella). 2. G. pectinata (G. plannlata). 3. G. incrustans. 4. G. mantonae, n. sp. 5. G. benhami Vaughan. 6. G. seychellensis. Of these it is barely possible that incrustans, mantonae, benhami, are varieties of pectinata. The mode of division of calices is rarely given by authors, but, judging from the speci- mens before me and the numerous figures published, the “ subequal division ” of Vaughan’s definition is the rule in all but G. mantonae. I tabulate the species as follows : a. No tendency to meander : (1) Multiplication by subequal fission. Calices up to 4 mm. wide . . . . . . 1 . G. retiformis. (2) Multiplication by marginal fission. Calices up to 10 mm. wide . . . . . . .2 . G. mantonae. b. Occasional short meanders : (1) Calices up to 7 mm. wide; meanders rarely over 12 mm. Palial lobes large . (2) Calices 10 mm. wide ; meanders up to 25 mm. long. Palial lobes poorly developed . . . .5. c. Meanders frequent, 6 to 9 mm. wide, 30 mm. to 44 mm. long 6. G. pectinata. G. incrustans. G. seychellensis. G. benhami. Goniastrea retiformis (Lamk.). 1914. G. retiformis Matthai, p. 118, pi. 31 figs. 1-5 ; pi. 33, fig 3 ; pi. 38, figs 2 and 4. 1914. G. solida Matthai, p. 117, pi 28, figs 3 and 4 ; pi. 31, fig. 1 ; pi. 33, fig. 4 ; pi. 38, fig. 3. 1918. G. retiformis Vaughan, p. 114 [pi. 15, fig. 24; pi. 16, fig. 25 ; both of Dr. Mayer’s article (1918.)] 1918. G. jparvistella Vaughan, p. 114, pi. 44, figs. 2, 2a. 1925. G. retiformis Hoffmeister, p. 26. * Crossland had added, “ and is described as such on p. — of this report.” I am unable to trace this reference. — [A.K.T.] VI, 3. 10 134 GREAT BARRIER REEF EXPEDITION The typical and commonest form, with its small, sharp-walled cahces, is one of the most easily recognizable of all corals, but its variation is extensive and gives puzzling cases. The history of the species is as follows: Matthai (1914, p. 118) includes in his synonymy-list 7 references under the name retiformis, 3 under solida and 5 under other names, which means that both Milne Edwards and Haime and Gardiner are unable to distinguish these two species. On the same page, but under G. solida, he says that of the five specimens which Milne Edwards and Haime refer to G. solida, two belong to G. retiformis. Of the 19 “ Pola ” specimens at Vienna, all referred to G. favus by Marenzeller, he says that ten are undoubtedly G. retiformis and that the remaining ones may be assigned to G. solida. No. 15918 (solida), Matthai says, has corallites showing a meandering tendency. By definition above this is not solida, but probably planulata. Also I consider that Matthai’s (1914) fig. 1 on pi. 30 is of G. solida according to the text on p. 117, but G. retiformis according to the text on p. 118, and the explanation of the plates on p. 139. In 1918 Vaughan says that the name solida is inappropriate, and substitutes G. parvistella Dana since Matthai took it from Eorskaal, whose Madrepora solida is a Porites, as discovered by Marenzeller. Since Forskaal says the species is the building material of considerable towns there was hardly need to wait for Marenzeller to know that it is a Porites ! Matthai, however, distinctly writes “ non Madrepora solida Forsk.,” so the objection does not hold. The above prepares us for Hoffmeister’s finding that, in his Samoan specimens, “ the main difference that I am able to make out in all these specimens is in the thickness of the walls. The Samoan specimens, however, show walls as thin as those in typical G. retiformis in some parts of the corallum, and, in other parts, are as thick as those of Dana’s type of G. parvistella .” “ I have examined very carefully all of the specimens in the U.S. National Museum, including those in the Samoan Collection, and have com- pared them with Dana’s type of G. parvistella and with Matthai’s description and figures of G. solida and G. retiformis.” Hoffmeister therefore concludes that the species are the same. Umbgrove (1939, p. 32) agrees, and “ will describe transitional forms from Togian reefs.” Finally in 1934 Matthai labelled two specimens of the Great Barrier Reef Collection, No. 142 ? G. solida, and No. B.M. 183 from Batt Reef as G. retiformis. The two are exactly alike, except that the latter is somewhat more solid. In both, pah may or may not be divided by a deep notch from the septa, the sides and edges of which are spinulose, very much so in the “ solida ” specimen, the columella degenerate and often fused. It is clear that this species is an excellent subject for ecological work on the reefs, the easier as the species is so common. For the present the species name is retiformis, with variety solida. The other two specimens, No. 226 T. 1 moat and No. 177, are typical, and are matched by numerous figures by several authors. No. 142 B.M. 185 is best matched by Matthai’s pi. 31, fig. 1 ; and No. B.M. 183 by Mayer’s (1918) pi. 15, fig. 24. S. Manton refers to this species on pp. 285, 286, 288, 293, 294, 305 and 307, and on plates viii, xi, xii, xiv and xvi. On Traverse I it is one of the most abundant corals in the moat ; at about 830 feet the rocks dip below low water, and this species appears with F. favus, F. astraei- formis and Acropora spp. On the seaward slope between 846 and 930 feet it is more abundant than the other Faviidae, which become more numerous as the water deepens. MADREPOR ARIA, HYDROCORALLIXAE, HELIOPORA. AND TUBIPORA— CROSSLAND 135 On the seaward slope, as on Traverse II. are the largest specimens, as well as most numerous species, up to 16 inches across — quite a large coral for Low Isles. PI. viii shows its extension from lowest tide level to 8 feet below, but it extends no further. On Yonge Reef it occurs both on the crest, pi. xiv showing one flat-topped colony 18 x 11 inches, and in the anchorage zone, where pi. xvi shows one 32 x 18 inches, thus showing that it is tolerant of very variable conditions. Only one specimen of the solida form is recorded, 3 inches in diameter on the seaward slope of Traverse II. It is abundant, and much larger in the Maldives. Gardiner (19046, p. 772) writes : This golden-green species is common on the outer slope, occasionally occurs on the reef- flat in small heads, and is abundant in the lagoon, both at Minikoi and in the Maldives. It often forms immense masses, which on the shallow flat behind the boulder zone die in the centre but spread linearly, the blocks in growth resembling massive Porites.” I have not seen this golden-green colour in the Red Sea, where it is very light brown or almost white. Distribution : Vaughan says, “ One of the best known Pacific and Indian Ocean corals.” Common, often abundant, from the northern Red Sea to Fiji and Samoa, but not in Tahaiti. 2. Goniastrea pectinata (Ehr.). 1892. Coeloria australiensis Rekberg, Abh. Xaturwiss. Vereia Hamborg, pp. 1-50, pis. i-iv. 1914. Goniastrea pectinata Matthai, p. 120, pi. 28, fig. 6 ; pi. 37, fig. 1. 1914. Goniastrea planulala Matthai, p. 121, pi. 28, fig. 5 ; pi. 31, figs. 7 and 8. 1918. G. pectinata and planalata Vaughan, pp. 114 and 116 ; pis. 42 and 43 ; [pi. 15 of Dr. Mayer’s article (1918)1- The two species recognized by both Matthai and Vaughan differ only in the thickness of their walls. The two works quoted may be consulted for the difficulties both these highly experienced men have had in separating the species, which I will not repeat in full, but, as one example, on p. 115 Vaughan says “ Dr. Mayer obtained 20 other specimens of Gonistrea pectinata on the Murray Island reef. The variation is simply bewildering . . . The principal variation is in the character of the wall between adjacent corallites and in the depth of the calices.” These characters vary with position on the colony in the usual way, but the means also vary from colony to colony. There is no difference in the present collection between some of the specimens labelled pectinata by Matthai, and those he has labelled planulata ; all the specimens can be arranged in a continuous series. Matthai labelled the specimens as follows : (1) As “ G. pectinata .” No. 224 0. T.l Moat ; and the following unnumbered from Batt Reef ; B.M. No. 171, Patch 1, square 4 ; B.M. No. 172, Patch 1, square ? ; B.M. No. 175, square 3 ; B.M. No. 176, Patch 1, square 9. The following are marked with a query : Nos. 153, 146 and B.M. 88, locality 2. These last three specimens have very thin walls, slightly larger and deeper calices, and narrow septa, but do not differ from No. 224 preceding, which is not so queried. I do not share Prof. Matthai’s hesitation. His identification of B.M. No. 88 with Coeloria australiensis is not queried; the ? is to Goniastrea sp. There is no difference between it and Nos. 153, 146 and others. (2) As " G. plamdata.’' From Batt Reef, Patch 1, square ? ; B.M. No. 178, a half dead scrap ; B.M. No. 179, similar ; B.M. No. 180, fragment of a small but living colony ; 136 GREAT BARRIER REEF EXPEDITION B.M. No. 181, showing some marginal fission. The last-mentioned is a whole colony, though very small (7 cm. in diameter x 7 cm. in height), and columnar in shape. Walls thin to very thin, calices low down the sides generally, hut not always, thicker walled and shallow. It combines the characters of G. planulata and G. pectinata. S. Manton refers to G. pectinata on pp. 284, 285, 287, 293, 294, 300 and 302, and on pis. iii and x. As she does not refer to G. planulata I presume it is included in G. pectinata. It is found in very small colonies, 1 to 4 inches across, in shallow parts of Traverse I, as far as 620 feet (see pi. iii, graph 22) and is one of the more abundant corals (1 to 7 inches across) in the moat, but is absent from deeper water. It occurs also on the inshore part of Traverse II, but is again absent from the seaward slope. It is found to remove sediment, both sand and mud, with relative ease, which probably accounts for its successful growth in shallow water. PI. X shows only two little growths, 4| inches across, in the moat. On this part of the Great Barrier the species is very much less common than is G. retiformis. Gardiner did not find it in the Pacific Islands he visited, and it is somewhat rare in the Maldives, though it is common at the Murray Islands and Samoa. Distribution : Red Sea, Indo-Pacific to Samoa, but not to Tahaiti ; and apparently absent from some islands west of Samoa. 4. Goniastrea benhami Vaughan. (Plate VIII, fig. 2.) 1917. Vaughan, p. 277, pi. xviii, figs. 1, 2, 2a ; pi. xix, figs. 1, la ; pi. xx, fig. 1. 1918. Vaughan, p. 116. One small specimen, No. 51, of this rare species is in the present collection, and another was brought to me by Mr. G. W. Otter. Vaughan published 6 figures in the Trans. New Zealand Inst. As this publication may not be always accessible, I publish another, but do not repeat Vaughan’s description except to quote “ except that it has meandroid calicinal valleys it bears a considerable resemblance to some specimens of G. pectinata .” I also remark that, in this specimen, valleys are meandroid in all parts of the colony. They are up to 30 mm. long, with 4 or 5 centres. Distribution : Kermadec Islands, Formosa, and now the Great Barrier Reef. Goniastrea mantonae, sp. n. = Goniastrea K.5 and K.18 Manton (1935). (Plate VII, figs. 1 and 2.) The only three fair-sized specimens of this species are here assembled. They are labelled by Prof. Matthai only as Goniastrea sp., their numbers being : sample 120, B.M. 186, T.2 shallow ( Goniastrea K.5, Manton) ; 128, B.M. 404, T.2 shallow ; 179, B.M. 187 (Goniastrea K.18, Manton). They were labelled by letter and number because Dr. Manton could not identify them with any known species. The first is apparently half a regular hemisphere, 12 cm. in diameter, 6 cm. thick. No. 128 is a whole colony, a crust 0 to 5 cm. thick, growing over some other Faviid. No. 179 is a nearly regular low dome, 12 cm. across by 5 cm. high. All were in good condition and active growth. UADREPORARIA, HYDR OCORXLLENAE, HELIOPORA AXD TUBIPOR A— CROSSLAND 137 I am fortunate in having three good specimens on which to found a new species, but, as in most coral genera, and especially in this, a dozen would have been none too many. Definition : As G. pectinata, but (1) calices are larger, the largest in Iv. 18 being 10 mm. x 9 mm.. 11 mm. x 7 nun., 12 mm. x 10 mm. ; in No. 128 the largest are 10 nun. X 7 mm., 9 mm. x 5 mm., 11 mm. x 7 mm. ; and in Iv5. are 9 mm. x 9 mm., 10 mm. X 9 nun., and 12 mm. x 7 nun. ; the depths are in proportion as in G. pectinata. (2) There is no trace of meanders. (3) Fission is always marginal. Palial lobes always distinct, but not thicker than the upper halves of the septa. 'While the latter are finely denticulate the former bear small teeth : notch between palial lobe and upper part of septum horizontal, or extends downwards a little way. Columella deep below palial lobes, of very fine, often indistinct, trabeculae. In No. 128 the intermediate septa are rudimentary and often absent ; in lv.5 small but generally present; in Iv.18 longer and always present. Edges of septa strongly denticulate in all three, sides roughly granulate in K.5 and No. 128, nearly smooth in K.18. K.5 has thicker walls and septa than the other two, and, as usual, the walls are still thicker and calices shallower near the base. The thicker-walled K.5 is shown on Plate VII, fig. 1. The general likeness to G. pectinata is obvious, but the marginal fission and absence of meanders are equally distinct. Owing to differences between the specimens all three must be regarded as syntypes.* They cannot be considered apart. S. Manton refers to this species on pp. 287, 288, 293, 294, 305 and 306, and on Plates viii, xi, xii, xiv, xv and xvi. On both Traverses I and II it is present in the moat and on the shallower part of the seaward slope, much as G. pectinata. On Yonge Reef both on the crest and on the inner part, where it is frequent. It thus endures a wide range of conditions, though it never extends much below low water. Manton’s plates also show a wide local distribution, though it is never abundant. Pis. xv and xvi show larger specimens, up to 34 cm. across, the examination of which would have made my definition of the species much more complete and safe. [Family Astraeidae : Meandroid species.] Genus Cynarina Briiggemann. 1877. Revision of the Recent Solitary Mussaceae. Annals and Mag. Nat. Hist., Ser. IV, XX, p. 305. A genus with only one species, known only from the Gulf of Suez, where it was found by Savigny and later by MacAndrew. Now a single specimen from the Great Barrier Reef. It is distinguished from Sclerophyllia margariticola Klz., also from the Red Sea, by its narrow base and thin septa, among other tilings. Cynarina savigny i Briigg. (Plate IV, figs. 1, 2.) 1826. Caryophyllia ccirduus Audouin, p. 233. 1827. (Caryophyllia carduus ) Savigny, pi. 4, fig. 2 (1-3). 1877. Cynarina savigny Briiggemann, p. 305. * Crossland wrote “ form the type.” In accordance with the International Rules B.M. Reg. No. 187 is here designated holotype, and Nos. 186 and 404 paratypes. [See also footnote, p. 125. — A.K.T.] 138 GREAT BARRIER REEF EXPEDITION The single specimen, No. B.M. 492, is the largest yet found, being 4| cm. over the long diameter of the calyx, and 5 cm. high, Briiggemann’s largest being 3-| cm. x 4 cm., a little larger than Savigny’s. Savigny’s figures are perfectly reliable, but Bruggemann remarks, as I have had to do, that his “ figure was either mistaken ” (I suppose misunderstood is meant) “ or overlooked by subsequent authors.” However, I refigure the larger specimen before me, as Bruggemann’ s is the only account between 1826 and the present day. I agree with Bruggemann in substituting savignyi for Audouin’s name carduus, since apparently that was given under the mistaken idea that the species is the young of a Lobophyllia, and carduus has been given to several species in this family, mostly badly described. The most remarkable features of the species are the thin septa ( cf . Sclerophyllia margariticola Klz. and Mussa ( Lithophyllia ) lacrymalis M. E. and H. which are probably the same) and the large columella made of closely packed thin lappets (see Plate VII, fig. 2). These features also make it impossible to be a young form of any species of Lobo- phyllia. True, in this specimen there are about 12 (many are broken) more exsert, more coarsely toothed septa, but these are not greatly thicker than the others, and almost all septa, even the thinnest and narrowest, reach and contribute to, the columella. Details can be obtained from the photograph. The toothing of the larger septa, especially their exsert parts, is very coarse — sinuous or lobed might better describe them. The smaller have more numerous, but still rounded teeth. In contrast with Savigny’s figures, and presumably Briiggemann’s specimens also, only one constriction is at all deep, but three are distinct, and two more above these can be seen faintly. The theca and epitheca are so reduced in the upper part that it is full of squarish holes between the septo-costae and dissepiments : these are to be attributed to age and have no morphological significance. Endothecal dissepiments make the cup very shallow, but, as they slope inwards, it is 5 mm. deep over the columella. Genus Lithophyllia. 1857. Milne Edwards and Haime, p. 290. 1877. Scolymia Bruggemann, p. 301. 1899. Lithophyllia Gardiner, p. 166. The adoption of Haime’s name Scolymia by Briiggeman in spite of Milne Edward’s and Haime’s statement that it had not been published, is unjustified. Lithophyllia vitiensis (Bruggemann). (Plate IX, fig. 4.) Scolyjnia vitiensis Bruggemann, l. c., p. 301. Lithophyllia vitiensis Gardiner, p. 166. One specimen corresponding very well with Briiggemann’s from Fiji, dredged at Stn. XXIV, B.M. Reg. No. 1934.5. 14. 410c. It has not been figured hitherto. It measures 36 mm. x 25 mm., the type being 40 mm. in larger diameter. The substratum is an extremely rotten fragment of shell with a nullipore on it, the growth of which quite obscures the base, but the corallum appears to be 7 mm. high on one side and 3 mm. on the other. It is thus much more squat as well as more oval than the type. MADREPORARIA. HYDROCORALLTXAE, HELIOPORA AND TUBIPORA— CROSSLAND 139 Briiggeman says " no distinct epitheca.” In this specimen it is distinct, but only like a thin, shrivelled-looking membrane, so thin that the low costae are clearly seen ; it reaches nearly to the top of the wall. Septal granulations not visible unless suitably lighted, when they become quite distinct. The oth cycle is incomplete, probably owing to the distortion of the specimen and its smaller size. The columella is peculiar, and as Briiggemann describes it, “ very dense, with subimbricate surface, the trabeculae being- enlarged to horizontal, somewhat crimpled lamellae.” I hope the figure on Plate IX will explain this somewhat peculiar arrangement. I add that the surfaces of the lamellae are dotted with small granules like those on the septa, and that the circular columella is quite small in comparison with the area of the quite flat theca. Recorded Distribution : From Fiji before 1877 ; Loyalty Islands by Willey before 1899 ; the only previous records. Genus Caulastrea. For the relationships of this genus, which has special morphological interest, see Matthai (1928), pp. 11, 13, 14, 17 and 272. Briefly it is a primitive member of the Mussa group, the Indo-Pacific representative of the Atlantic Protomussa. Caulastrea simplex sp. n. (Plate III, figs. 4, 4a.) A single specimen, dredged from Stn. XVI, of a small solitary coral, 12 mm. high and 8 mm. in diameter, belongs to this genus. It is not a young specimen of any compound species, since the eversion of most of the septa over the top of the calyx indicates that it is fully grown. Its form is turbinolid, with a bent base, but this may be accidental, and due to the shape of the shell fragments to which it is attached. The upper part is nearly cylindrical, capped by the large exsertions of the septa, which exsertions add 2 mm. to the height given above. These eversions are continued to the base as distinct but narrow costae, all of nearly the same size, widely apart. They correspond to the first three cycles of septa, and bear, like them, triangular teeth, but much smaller and more rarely. The calyx opening is quite circular, and 3 mm. deep to the columella, i.e., 5 mm. from the top of the larger septa. The septal series are consequently regular, the first two alike, the third narrower, especially below, where it does not reach the columella, and the fourth is rudi- mentary. The septal exsertions and their teeth are more like those of e.g., Lobophyllia than those of the other species, the teeth being broad and blunt, the largest at the outer angle of the exsertion ; but the teeth of the costae are small and far apart. These are shown on Plate III, fig. 5, but those of the interior of the calyx cannot be well shown. As the calyx is deep and the septa are narrow, the greater part of their edge is nearly vertical ; this part bears three or four broad blunt teeth, narrower and sharper on the 3rd order. All these teeth, like the septal edge and exsertions, are granular, the sides being smooth until examined in a carefully adjusted light. The outgrowths which form the small and rather loose columella are smooth and flattened. The differences between this and the other species of the genus are considerable enough, and the fact that this is a solitary coral might be taken as reason for the creation 140 GREAT BARRIER REEF EXPEDITION of a new genus. This is not necessary, however, since the differences are of degree rather than kind, and the solitary habit is not, of itself, a generic character. Caulastrea simplex is also not unlike Parasmilia centralis M. E. and H. from the upper chalk (1850, p. 47, pi. viii, figs. 1, 1 a-c). It differs mainly in the toothing of the septa, those of Parasmilia being entire as in other Eusmiliidae. Caulastrea furcata Dana. 1928. Matthai, p. 273 ; pi. 44, fig. 5 b and 6 ; pi. 45, fig. 3 ; pi. 61, fig. 3 ; pi. 62, figs. 6 and 12. This species was dredged from Stn. XVIII and XXI. There is no particular difference between the sets of specimens, or between them and Matthai’s complete description and figures, except that these are more loosely branched, and the edge-zone extends well down the stems. Distribution : Already known from the Great Barrier Reef, but other records are confined to the Pacific, between the China Sea, Fiji and Tongatabu. Genus Acanthastrea M. E. and H. 1914. Matthai, Favia, p. 77. 1918. Vaughan, Acanthastrea, p. 101, p. 125, pis. 50, 51. 1924. Matthai, p. 37, pi. 3, figs. 2 and 3 ; pi. 5, fig. 1. 1931. Crossland [Favia Jiemprichii], p. 387, pi. 21. The genus is defined by its possession of teeth or spines on the upper parts of the septa, especially over the walls. This definition is insufficient without the distinction between, e.g., teeth which are the result of reduced growth and those due to added growth. As more pertinent examples than those I gave in 1931 of the former, I cite those on the upper edges of interthecal costae in Favia dipsacea, Aud. and Sav., rediscovered on the coast of Natal by Prof. T. A. Stephenson and his colleagues. Their existence, combined I suppose with distrust of Savigny’s very excellent figure, has caused this species to be included in this genus. Similarly F. parvimurata Gard. is given as a possible synonym by Vaughan, and definitely by Thiel, though the teeth on its exsert septa are of the most ordinary sort. (See illustrations and references under Favites aspera on p. 132.) The spines of Acanthastrea are very different, and quite a special growth, with triangular bases and thinner prolon- gations with spinulose ends. Many of them are hollow, and all the longer ones are removed by handling before the specimens reach museums, leaving traces, when hollow, as little holes that appear in illustrations as small black dots. To show a complete specimen, with the fullest development of the spines, I had one photographed in Tahaiti, which is shown in the above quoted paper. The relationships of the genus have been somewhat confused. Matthai in 1914 brings the two species he describes under Favia. Vaughan in 1918 separates Matthai’s species into Acanthastrea , which, from the position in which he places the species he describes, he evidently regards as far removed from Favia. Favia complanata, superficially resem- bling an Acanthastrea until its spines are examined, he rightly places in Favites, and near to F. abdita. The septa being the foundation of an Astraeid skeleton, from which both theca and columella are directly derived, I follow Vaughan in attaching great importance to their characters. I therefore, like Vaughan, regard this genus as quite distinct from Favia MADREPORARIA, HYDROCOR A LLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAXD 141 and allied to Lobophyllia (Mussa) and beg the reader to refer to Vaughan (1918). pi. 50, on which figs. 16 and 2 bring out the relationship in a striking way. In 1924 Matthai, pi. 5, fig. 1, and pi. 3, figs. 2 and 3, shows a specimen of a species with huge calices, which, he names " Acanthastrea or Mussa sp.,” which is certainly of this genus, differing mainly from A. echinata in the size of its calices, the formation of short valleys being a common occurrence in the smaller species (Matthai, 1914, p. Ill ; Crossland, 1931, p. 287.) My Symphyllia simplex from Natal is similar, but not identical, and might equally well have been placed in Acanthastrea . Acanthastrea echinata Dana, 1840. (Plate VIII, figs. 1 and 3 ; Plate IX, figs. 1, 2.) 1879«. Prionastrea spinosa Klunzinger. p. 39, taf. 4, fig. 7. 1914. Faria hemprichii Matthai, p. 110, pi. 27, figs. 1, 2 ; pi. 36, fig. 3. 1914. Faria hirsuta Matthai. p. 100, pi. 24, figs. 7, 8. 1918. Acanthastrea echinata Vaughan, p. 125, pis. 50 and 51. 1931. Faria hemprichii Crossland, p. 387, pi. 21. The name hemprichii is due to Ehrenberg, whose description is nul* ; then to Milne Edwards and Haime, who describe as Ehrenberg's something which is both unidentifiable and not an Acanthastrea ; and finally Matthai in 1914 publishes an unsatisfactory photo- graph of Ehrenberg’s type, with which he identifies his species “ with some hesitation.” To me this hesitation amounts to extreme doubt, and I therefore regard hemprichii as a nomen nudum still. I regard both Matthai’s names as the same, and take Dana’s Acan- thastrea echinata as the earliest name available. Fortunately, Vaughan has made this safe by his description and figures of Dana’s type, without which it was confused with any Faviid having teeth on the exsertions of the septa. I have carefully tabulated the characters given by Matthai for his two species, and find that there is no valid distinction. The two specimens before me are illustrated on Plate VIII, figs. 1 , 3, and are seen to differ very greatly in appearance, but this is almost entirely due to the differing thicknesses of the walls, with which go polygonal calices in the one, and almost round in the other, but it is a commonplace that such differences are not specific. In both specimens the septa are thin ; the characteristic thickening over the walls is but slight in No. 151, fairly well marked in No. 144, but in neither so strongly marked as in 1, e.g., Dana’s type. The columella is generally neglected in descriptions, and I am uncertain what weight should be given to its details. Such figures of this species as show it appear to give an ordinary trabecular arrangement with simple teeth, Matthai says, “ pointing upwards ” ; neither of these specimens is of this type. In No. 114 they are of the F. abdita type, figured, for F . virens, on Plate VI. But in No. 151 they are quite peculiar, made of greatly thickened, blunt, smooth septal teeth pointing level to the centre. As the uppermost of these highly modified teeth do not reach the centre, the columella is hollow on the top. A very few, minute, spinulose projections do occur in some calices, so that possibly the structure is not so far removed from that of No. 144 as seems at first to be the case. How- ever I illustrate both on Plates VIII and IX. * Dr. Crossland’s meaning is not clear. At the bottom of p. 96 (Ehrenberg, 1834) appears : “ 9. A. Hemprichii E.” At the top of p. 97 is a 6-line description. Perhaps Crossland missed the latter. — [A.K.T.] 142 GREAT BARRIER REEF EXPEDITION No. 151, “ hirsute/,” is very well matched by Klunzinger’s taf. 4, fig. 7, as Prionastrea spinosa (which Matthai makes the same as F. hemprichii). The columella is visible, with a lens, in two calices ; it is of thick beams, probably like this Great Barrier specimen. The only other figures at all like our specimen are Matthai’s (1914) pi. 27, fig. 1 (F. Jiemprichii), and Vaughan’s (1918) pi. 51, fig. 2 (A. echinata), from Murray Island. As for No. 144, there is nothing quite like it in the literature, and I therefore illustrate it on Plate IX, fig. 2. Distribution : From the northern Red Sea, through the Indian Ocean as far as Tahaiti. Always rare. Genus Lobophyllia. Stephenson and others mention the genus on pp. 67, 86 and 88 as one of the animals characteristic of the seaward slopes and anchorage at Low Isles, of Yonge Reef, and of the reef patch at Lizard Island. S. Manton refers to the genus on pp. 289, 295, 298 and 302, and on pis. vii, viii, xi, xii and xiii. On Traverse 1 the genus is met with in colonies of 2 to 12 in. occasionally in the moat, and on the seaward slope. On the windward traverse it occurs only rarely between “ distances ” 700 and 760 feet in water 8 to 10 feet deep. The plates of drawings show only rare and insignificant scraps. This is in very great contrast to the Red Sea, where, in both northern and central sections, numerous colonies of L. corymbosa make masses up to 10 feet high and 20 across ; but the other two species, so far as I know, are smaller, though not at all uncommon. (Crossland, 1935, p. 502, and 1939, p. 515.) The Great Barrier conditions recall those of Tahaiti as far as this species is concerned. The only other reference to “ enormous colonies ” is by Umbgrove (1939, p. 37) quoting Vervey, in the Bay of Batavia. The variation of the three species was worked out by me in Tahaiti (1931, pp. 373 to 380, and table on pi. 22), and though it is possible that variation is at a maximum in that far oceanic island, the present collection, like all the literature, shows an astounding amount of variation. Lobophyllia corymbosa (Forsk.). (Plate IX, fig. 3.) For synonymy see Matthai (1928), p. 210, and pis. 25-27, 47,* 57, 58, 60, 62, 64, 68, 71. The species is the only one of the three which is certainly definable. The collection contains one specimen and 6 fragments. No. 110 is a regular dome 20 cm. in diameter. It resembles Matthai’s (1928) pi. 68, fig. 1, “side view of Dana’s type of Mussa cactus ,” the calices of which are shown on pi. 26, fig. 4, which is very near to Forskaal’s type, pi. 71, figs. 5 and 6. Dana’s type seems to be about the same size as No. 110, and apparently is the largest specimen to which Matthai had access, though he saw 106 specimens at the British Museum (Natural History), and 30 abroad, of which only 50 had a maximum diameter of 20 cm. No. 447 is remarkable for the thickness and exsertion of its septa, up to 6 mm., including the teeth. The upright teeth on the exsert parts of the septa often number three to five, * PL 47, fig. 8, is named Lobophyllia costata Dana. Crossland may have included it in error.— [A.K.T.] MADREPORARLA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 143 and the upper costal teeth are unusually prominent in this, unlike any illustration I know of this species, except Crossland’s (1931) pi. xii, fig. 23. It is near to Matthai’s (1928) pi. xxv, fig. 6, “ probably Milne Edwards' and Haime's type of Mussa fistulosa,” or to pi. 27, fig. 1, “ probably Milne Edwards’ and Haime’s synfype of Mussa aspera .” It is more like the former in that the septal teeth, where broken, are seen to be hollow. The columella is very small in Xo. 447, as in both these type-specimens. As I wish to show the septal teeth and " eversion ” of the exsert parts of the septa the illustration on Plate IX is taken a little oblicpiely. Distribution : Xortliern lied Sea, through the Indian Ocean, including Mauritius and llodriguez, and Pacific out as far as Tahaiti. Lobophiyllia hemprichii (Ehr.). (Plate X, figs. 1 and 2 ; Plate XXX, figs. 1 and 2.) Two specimens are labelled L. costata by Matthai, viz., Xo. 2oo, a half dead scrap, and Xo. 94, a well-preserved specimen 19 cm. across and 10 cm. high. A third specimen, B.M. 380 " “ Lobophyllici ? Stn. Xo. 2 Low Island ” is very like Xo. 94, but is only part of a colony ; its differences will be given later. All three specimens are alike in the breadth of continuous valleys, and in the remarkable long, tapering teeth of the exsert parts of the septa. There are usually three of nearly ecpial length (5 mm. up to 8 mm. ; one is vertical, the other two point inwards and outwards at an angle (Plate X, fig. 2). The fourth specimen, Xo. 450, is labelled L. hemprichii , with which I agree. I cannot, however, agree that any of the preceding correspond with Matthai’s definition of costata, since the vide valleys, 25 to 30 mm., are continuous. In Xo. 94 they are 15 to 20 mm. deep ; in B.M. 380 shallower, but 13 to 17 mm. I have described the variation of the species costata and hemprichii in Tahaiti (1931, p. 379, table on pi. 22), showing a difference between the species in spite of overlapping. B.M. 380 and Xo. 94 differ greatly in the parts of the septa within the calices. In Xo. 94 there are 4 main septa per cm., of which one is generally thin, narrow, with fine teeth, and only twro large teeth above the wall. The tertiary rudimentary, very finely toothed ; intermediates alternate regularly, so adding 4 to the total number of septa per cm. In B.M. 380 there are 34 large septa per cm., all alike, and the intermediates are quite rudimentary ; the lowrer teeth of the main septa are only 1 to 3, small and triangular in this specimen, but up to 7 small sharp teeth in Xo. 94, but parts of Xo. 94 approach B.M. 380 in condition of the larger septa, and Xo. 255 is intermediate. The following published figures of Matthai (1928) match these two specimens : pi. 28, fig. 6. A specimen from Torres Straits ; its valleys are narrower and teeth shorter. Pi. 29, fig. 4, Milne Edwards’ and Haime’s Mussa echinata , is similar, but its valleys are only 20 mm. wide, and apparently all the longest teeth are broken off. PI. 66, fig. 2, Ehrenberg’s type of Manicina hemprichii, valleys 25 to 30 mm., wide and teeth occasionally prominent. All the above are named L. hemprichii by Matthai. The specimens also match Faustino’s (1927, pi. 39, fig. 2, Mussa cristata Ehr.). Milne Edwrards and Haime, in describing the species specially mention these peculiar teeth, but they are blunter than in our specimens. This specimen is called costata by Matthai. As these teeth are not well shown in any of these pictures, and I feel that I must offer evidence 144 GREAT BARRIER REEF EXPEDITION in support of my disagreement with. Matthai, I give illustrations on Plate X, figs. 1 and 2. In order to show the septal teeth the views are rather oblique. No. 450 is entirely different in the structure of the exsert parts of the septa, and therefore differs from the first three in appearance as much as any two distinct species could. The columellae also are larger and more closely compacted, and are joined by from two to five lamellae. The tips of almost all the upper teeth are broken off,* but they seem to be only elongated triangles, not spikes, as in the preceding. They are hollow. Of these comparatively small and uniform teeth (Plate X, fig. 1) there are five on the upper edge of the septum, but as this edge forms a regular rounded arch, with no boundaries between inner, upper and costal parts, the number cannot be counted exactly. At the inner angle, or a little below it, is an extra large tooth, and below it one to three rather smaller ; on septa of the second order these are sometimes followed by three or four small thin sharp teeth. Septa of the third order are very small, but bear teeth above, and below are sinuous, or bear one or two small teeth low down. This form somewhat resembles Yabe, Sugiyama and Eguchi’s (1936) pi. 32, fig. 1 ( L . hemprichii), but the differences are considerable enough to warrant another illustration. Distribution : Red Sea, Pacific to Tahaiti, but the only Indian Ocean record is a doubtful one by Gravier from the Gulf of Aden. Genus Symphyllia. Stephenson and others refer to the genus on pp. 67, 80, 86, 88 and 90 as one of those characteristic of the seaward slope and anchorage at Low Isles ; at Three Isles it forms, with Favia and Pontes , flat-topped platforms in a pool between the mangrove swamp and shingle spit ; it is found also on Yonge Reef and the reef patch at Lizard Island and on Batt Reef. It is thus widely distributed, but never seems to have been conspicuous. S. Manton refers to the genus on pp. 284, 285 and 302, and on plates iii, vii, ix, xii and xiii . Small specimens, 3-8 inches, are found in the deeper parts of the moat, on the inner side of the seaward slope of Traverses I and II (Graphs 28 and 56 on pis. iii and vii). On p. 302 the genus is mentioned as being one of those which, having large polyps, can easily remove both mud and sand. PI. ix shows two very small growths in the western moat, but on the other plates examples of from 23 cm. to 32 cm. in diameter are shown. The genus (or species — there is only one common) is common and widely distributed at Low Isles, but is not flourishing, which may be connected with the lightness of the specimens collected. Distribution : Indian and Pacific Oceans, as far east as Samoa ; not found in the Red Sea (unless Klunzinger’s Isophyllia erythraea is a simple form of this genus), nor in Tahaiti. Symphyllia recta Dana. (Plate XI, figs. 2 and 3.) Symphyllia sinuosa Q. and G. All authors up to Vaughan (1918). 1918. S. nobilis Dana ; Vaughan, p. 124. * Collectors should remember that, though stony and heavy, corals are delicate objects and need as careful handling and packing as other specimens. Most collectors should also remember that they do not know which specimens are of value and which are good for nothing but road material, but for eco- logical purposes all are of equal value. HADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 145 1918. 5. nobilis Mayer, pi. 17, fig. 35. 1924. S. sinuosa Matthai, p. 31. 1928. S. recta Dana. Matthai, p. 227 ; pi. 30, figs. 1-6 ; pi. 31, figs. 1,2; pi. 48, figs. 4-6 ; pi. 57, figs. la-5. 1932 to 1936. Yabe, Sugiyama and Eguchi, alternate between S. nobilis and S. recta. Synonymy : This species has been known as S. sinuosa by all authors up to 1918, and at least Milne Edwards and Haime’s description and figure were adequate as far as they went. It is very unfortunate that Dana’s greatly damaged and now quite worthless specimens were preserved and used as a basis for the resuscitation of his names nobilis and recta. The result is not only that a well-known name is altered, with no advantage to anyone, but certain loss, and even experts like Vaughan, Matthai and Yabe, Sugiyama and Eguchi cannot agree which name to use, but also that, instead of the identification of later specimens depending upon Dana’s types it is those types which are only recognizable from later specimens and the work done upon them. Dana’s samples should have been ignored. What Ellis and Solander and Lamarck meant by their descriptions is of no interest to any practical zoologist now. It is only by assuming that our present classification cannot be upset by the ecological research which is so badly needed that Dana’s names and types can have any validity. The following specimens are labelled S. recta by Matthai, viz., Nos. 13, 95, 112 and 209. Three specimens, B.M. 365, B.M. 367 and B.M. 405, are not named, except “ Sym- phyllia gonad species.” All are of the “ thin variety ” of Matthai except No. 209, which approaches the “ thick variety ” in having its main septa as much as 1 mm. thick. The three “ gonad species ” examples have particularly thin walls and septa, which latter therefore appear much less crowded than in the others, though there are 8 to the cm. in both cases. No. 13 is interesting as being a young specimen, 5 cm. in diameter, 4 cm. high, attached by its whole under-surface and with no radial arrangement of the valleys, and so distinct from S. radians. No. 95 is small and somewhat deformed. It has several monostomodaeal calices, most of the others being di- and tri-stomodaeal. Its septa are 0-75 mm. thick, in places 1 mm., and therefore appear crowded, though there are only 8 or 9 to 1 cm., as usual. The larger teeth have been broken off, but appear to have been short, giving the walls an evenly rounded appearance. There are no grooves. The following figures match these specimens ; No. 97, Matthai (1928), pi. 30, fig. 4, “ from Chagos ” ; Bedot (1907), pi. 21. No. 112 and B.M. 367, Matthai (1928), pi. 30, fig. 1, of Q. and G’s. type ; Bedot (1907) pi. 22 (but the inner septal teeth are less prominent). No. 209 is the only example of the thick variety, in which I place it because of its wider valleys and having three or four smaller septa between the larger, instead of the simple alternation of the other samples, and its large triangular irregular teeth, but in spite of the fact that its teeth are only 1 mm. thick instead of the 2-2*5 mm. of the definition. It is very like Matthai’s (1928) pi. 30, fig. 5, but as it is part of a larger colony its valleys appear longer. This figure is labelled “ thick variety,” though its septa are only 1 mm. thick. Gardiner’s (1899a) pi. 48, fig. 1, also affords an illustration. Nos. B.M. 405 and B.M. 365 have thinner walls and septa than any illustration pub- lished, but they grade in with B.M. 367 and No. 112, of which figures are given by Matthai 146 GREAT BARRIER REEF EXPEDITION and Bedot. Their upwardly pointed triangular teeth within the calyx are prominent, much larger than those on the walls, columellae well developed, compact, of thin rather broad lappets as in Lobophyllia ; septa 8 or 9 to the cm., larger and smaller alternating. I give a rather oblique view on Plate XI, figs. 2, 3, to show these features. Generally these thin forms are the less abundant. Gardiner’s Maldivan and Rotuman examples are all thick, the former, of which three good illustrations are given, having the main septa 2 mm. and over in thickness. He is the only author, except those of the Great Barrier Reef Expedition Reports, who gives notes on size and habitat ; as the contrast with Low Isles is marked I quote him. Writing (1899, p. 739) of Rotuma, he says : “ The species in the living condition is of a green colour, brown round the peristome. The colonies form great hemispherical masses, 2-3 feet across, and are very common in the outer half of the boat-channel. The species is noticeably resistant to the action of the sun, parts of the colonies at spring tides being uncovered for 2-3 hours. Massive colonies, too, which have died in the centre and been hollowed out, are rare.” Again (19046, p. 761) he says : “ The species is fairly common on the lagoon shoals and reefs of the Maldives. I have also seen it on the outer slope, and it is not improbably in places an important reef-builder . . . Colour of the Minikoi specimen ... a large mass several feet in diameter, in central upper part of the colony, over walls transparent, peri- stome very light green dotted with dark green, edge of stomodoeum white, and at sides, over walls very dark green, peristome white slate, edge of stomodoeum white.” Distribution : Indo-Pacific as far east as Samoa, whence Mayer obtained one small specimen. It does not occur in the Red Sea or in Tahaiti. Symphyllia radians M. E. and H. One specimen, labelled “ Low Isles W. moat,” B.M. 366, is quite typical in having a central stalk and radial valleys, but the septa are divided by grooves over the walls, generally just visible, but in two places 2 mm. wide. Principal septa 4 to 1 cm., as in Matthai’s definition, but there are generally three principal marginal teeth, often four, and the columellae are compact, made of thin broad lamellae, usually rounded. Matthai (1928) gives 5 figures, of which pi. 54, fig. 7, from Tizard bank, resembles this specimen fairly closely. PI. 58, the underside of Dana’s Mussa crispa, shows a central stalk with the rest completely free, but in this specimen growths and secondary attach- ments hide perhaps a third of the surface. In Bedot’s specimen, pi. 19, attachments seem to have covered most of the underside, and it is otherwise unlike the Gt. Barrier example. Distribution : Indian Ocean, but not east of Rotuma and Tongatabu in the Pacific. Rather strangely there are already four records from the Great Barrier Reef. Genus Oulophyllia. The two species of this genus were placed in C oeloria by Gardiner in 1904, an arrange- ment with which I have every sympathy, the present tendency to make numbers of genera with one to three species each being very inconvenient. Of the differences from Coeloria given by Matthai (1928), p. 256, the first, wider and deeper valleys, and second, septa narrower and thinner, are not generic ; while the third, a tendency to form palial lobes, is quite rudimentary ; and the fourth, distinctness of the columellar centres, is not complete, septal trabeculae occurring frequently between centres. As Matthai remarks, description MADREPORARIA, HYDROCORALLIXAE. HELIOPORA AND TURTPORA— CROSSLAND U7 of the polyps is needed, and it is very likely that the structure of the living and expanded polyps alone would settle the relationships of this genus to both Coelorici and Tridaeophyllia . Oulophyllia crispa (Lamarck). 1904. Coeloria cooperi Gard., p. 762, pi. lx, fig. 9. 1928. Oulophyllia crispa Mattliai, p. 257 ; pi. 19, figs. 1, 2 ; pi. 25, fig. 2 ; pi. 71, figs. 1 and 3. The first adequate figure is that by Gardiner in 1904. Rehberg’s figure of his U. maxima is only a drawing, not very distinct. His types are photographed by Matthai (1928), pi. 71, figs. 1 and 3. The present collection contains only one specimen, No. 52, a regular dome 12 cm. across x 7 cm. high. Distribution : Gulf of Aden, E. and W. of Indian Ocean, Singapore to Palau and Caroline Islands, and Bismarck Archipelago. Now Great Barrier Beef. Genus Coeloria. From S. Manton and Stephenson and others, we find that the genus is common on the Great Barrier, but inconspicuous. It appears, too, that C. daedalea is much more common than C. lamellina. Compare Gardiner’s (1899a, p. 740) description of its occur- rence in the South Pacific : “ The genus is very abundant on the lagoon-shoals at Funafuti, where it forms large, spreading masses, which vary in colour from brown to green. It is also found sparingly on the leeward reefs at Funafuti and Rotuma, but it cannot apparently withstand the force of heavy breakers." Of C. daedalea he writes : “ Common in the boat- channel, where it forms large spreading masses, 3-5 feet in diameter, small colonies only being found on the reef.” In the Maldives, says Gardiner (19046, p. 761) “ the genus is the most abundant of all Astraeids on the reef-flat and outside the edge of the reef to 5/., where the immediate force of the breakers is felt. It also occurs abundantly on the outer slope down to 15/., and one specimen comes from 28/. . . . Colour, generally some shade of green.” It is very remarkable that the above notes refer only to C. daedalea , C. lamellina being absent. In Gardiner’s collections from the central part of the Indian Ocean, Matthai (1928) records only two specimens from Chagos. Coeloria astraeiformis (M. E. and H.). 1879a. Coeloria esperi Kliinzinger, p. 19, pi. 2, fig. 6. 1899a. Coeloria astrceiformis Gardiner, p. 743, pi. 46, fig. 4. 1918. Mceandra astreiformis Vaughan, p. 120. 1918. Mceandra astreiformis Mayer, pi. 14, fig. 19. 1925. Maeandra esperi Hoffmeister, p. 29. 1928. Favia astraeiformis Matthai, p. 278 ; pi. 44, figs. 2a, 2 b ; pi. 45, fig. 1. One specimen, No. 230, Tl moat, names Favia astraeiformis by Matthai, while Nos. 143 and 148 are given a ? mark. Whether this mark refers to the genus or the species is not indicated ; probably to both, as the distinction from some specimens which Prof. Matthai has labelled C. daedalea , Nos. 168 and B.M. 78, are, if any, very slight. All authors but Matthai place this species in either Coeloria or Maeandra, which means simply that the gradation with Favia is complete. I therefore follow the majority. As regards C. esperi, Gardiner (1899a) thinks it very probably the same as C. daedalea, 148 GREAT BARRIER REEF EXPEDITION and Matthai in 1928 was definitely of this opinion. Vaughan (1918), p. 120, says it must be the same species as C. astreiformis, but Hoffmeister’s (1925) Samoan specimen of C. esperi “ agrees with Klunzinger’s (1879a) description and figure (p. 19, pi. 2, fig. 6) very well,” and therefore cannot be C. astraeiformis as Klunzinger’s species is the same as C. daedalea. The species is rare. S. Manton refers to it on 6 pages and 5 plates, but only once refers to C. daedalea. As a large number of the latter species was collected I assume that this means that the two cannot be distinguished on the reef. Distribution : Milne Edwards and Haime record it from the Eed Sea, where neither Klunzinger nor I have seen it. It has been found in widely separated places, in the Indian Ocean only by Gardiner. As far east as Fiji ; it is not recorded from Samoa. Coeloria daedalea (Ell. and Sol.). (Plate XI, fig. 1 ; Plate XII, fig. 2.) 1936. Coeloria rustica Dana, Wells, p. 104. The fact that Forskaal’s Madrepora daedalea is an Alveopora has given the opportunity to Wells to change the well-established name C. daedalea to Dana’s C. rustica. He is followed, quite uncritically, by Yabe, Sugiyama and Eguchi in 1936, and by Umbgrove in 1939* Matthai (1928, p. 24) gives 33 references between the years 1786 and 1927 almost all using C. daedalea for this well-known coral. Further references extend to 1935, making the period of unanimity 148 years. Wells’ proposed change cannot be too strongly repu- diated. (1) It is a defiance of the rules of nomenclature, which state that no change is to be made if greater confusion than uniformity results. Does Prof. Wells really wish to save us from confusing a species of Coeloria with one of Alveopora ? (2) Such changes bring the rules into contempt. (3) It depreciates the work of men who have made real additions to our knowledge of the species in favour of two bare records of local distribution. (4) It assumes that zoological nomenclature is the affair of a few museum experts only. I may also point out that “ Madrepora,” as Forskaal uses it, is not a genus but a rough division of the corals, as he himself definitely says. Of the 12 specimens from the Great Barrier seven deserve notice, viz., the two already mentioned, Nos. 168 and B.M. 78, which are not really distinguishable from C. astraeiformis; three with remarkably thick walls, one with a lamellar columella, and one with remarkably shallow valleys. The remainder, Nos. 140, 147, 176, 219 and 321, fall within the normal variation of the species given, by Matthai (1928, pp. 24-37). Specimen No. 147, which is 17 cm. long and 10-5 cm. high, consists of two areas with the usual meanders and sharp walls, the longest valley 28 mm. in length, most shorter, and between them an area 8 cm. x 5 cm. on which the walls are still thinner and the calices mostly single, as in C. astraeiformis. Another similar but smaller area is near one end. They are but slightly depressed. They are on the side of the corallum, the top of which is dead in the centre. * Dana’s type is figured by Matthai (1928), pi. vi, figs. 7 and 8. It is a beach pebble, so worn that even such authorities as Matthai and Vaughan do not agree whether it represents the present species or C. lamellina (Matthai, 1928, p. 29). It is therefore absurd, whether legal or not, to make such a defective specimen the type of an important species. MADREPORARIA, HYDROCOEALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 149 The longest valley is in No. 140, and is 50 nun. long, measured in a straight line. The specimens have been named by Matthai, but No. 54 with a ? mark, which it owes to its shallow valleys, 1-2 nun, deep and 4 mm. vide, and rather broad, rather thick septa, many of which slope inwards before dropping nearly vertically to the narrow columella, features which give the colony a very distinctive appearance. The specimen is a living area of a dead hemisphere, the dead parts partly overgrown by nnllipore and filamentous weed. Most of the dead area has been broken away, but in what remains along one side can be made out the higher thinner walls and narrower septa of more normal form. There is nothing like it in Matthai’s 18 figures, but, as it is clearly morbid, I do not illustrate it. No. 138 is a thin crust growing over dead Porites (?) illustrated on Plate XII, fig. 1. It is somewhat similar to the last but has thicker walls, but though the valleys are equally shallow, T75 mm. x 4 mm. wide, they are more open at the bottom. The most interesting peculiarity is the almost plate-like columella, like the figure Matthai (1928) gives on pi. vi fig. 1, “ showing variation towards the Platygyra facies,” but there is no reference to this remarkable fact in the text, under either genus. Gardiner (19045, p. 762) writes (under C. sinensis, which is most probably a synonym of C. daedalea) : “ When a valley is very shallow, owing to boring organisms underneath or other causes, the columella may approach in appearance the condition in Leptoria . . .” This specimen is free from parasites, but whatever other causes have induced growth as a crust about 8 mm. thick, a form strange to this genus, it does indicate the close relationship to Leptoria ( Platygyra ). Such a columella is also shown by Yabe, Sugiyama and Eguchi’s (1936) pi. xxi, fig. 10, of a curious humpy growth of “ C. rustica ,” but, as usual with these authors, the phenomenon receives no notice in the text. No. 321 is similar, and has been labelled “ Platygyra B ” by the collector, but the columella has nowhere the regularity of typical Leptoria. The thick-walled specimens also are not figured by Matthai or other authors, nor is anything like them found in his table of variations on pp. 34 and 35, all the specimens considered having sharp-edged walls, generally perforated. The valleys are deep and open. Their peculiarities are due to morbid deposits of endotheca, which not only swell the walls but make irregular floorings among the very loose columellar trabeculae ; or these may be absent, or represented only by long straight septal teeth. In the small scrap, No. 175, where walls are more normal the morbid deposition is less. It is similar to that found by me in Tahaitian specimens of Acanthastrea echinata (or Favia hemprichii) (1931, p. 387). Distribution : Many records from the Indian Ocean and Pacific as far as Samoa. It does not occur in Tahaiti, but Agassiz collected a specimen in the Paumotu Atolls, further east ( cf . Favia favus above). In the Red Sea it is rare ; it was collected by Klun- zinger (C. esperi) and Milne Edwards and Haime record it. The only specimen I have seen is semi-fossil, at Ghardaqa. Contrast C. lamellina. Coeloria lamellina (Ell. and Sol.). There are four specimens showing the usual great variation, numbered 61, 259, 261 and 371. Stephenson and others only refer to the presence of the genus in the fauna ; it is found everywhere, but the coralla noted by S. Manton are only 2-9 inches across, except one shown on pi. xv, the outer moat of Yonge Reef, which is 66 x 45 cm., but is a mere flat cake growing in only a foot of water. The big conspicuous domes and cylinders of the Red Sea are nowhere noted, vi, 3. 11 150 GREAT BARRIER REEF EXPEDITION No. 61 is a peculiar growth, 19 x 13 x 10 cm. high, which stood out bracket fashion from a vertical face to which it was attached by a narrow stem of ? Pontes ; the top is killed by exposure at low tide. The sides, both sloping and vertical, show long, nearly straight valleys with bulging areas at the edge with the usual meanders. On the sloping or outer side are thicker rounded walls and somewhat lamellar columellae. The colour of this species in the Red Sea is a particularly bright brown. Distribution : Abundant in the Red Sea, from which Matthai had 29 specimens, where it forms colonies up to 2 metres each way. Elsewhere it seems to be small and rare, especially in the Indian Ocean, but seems to be more common in the Pacific, where it extends to Samoa and Fanning Island. Genus Leptoria. 1834. Maeandra ( Platygyra , pars) Ehrenberg, non Platygyra Vaughan, 1901 ; nec Platygyra Briigge- mann, 1879. 1848-1857. Leptoria Milne Edwards and Haime. 1879a. Leptoria Klunzinger, p. 13. 1899a. Leptoria Gardiner, p. 739. 19046. Leptoria Gardiner, p. 764. 1918. Leptoria Vaughan, p. 117. 1924. Leptoria Matthai, p. 25. 1925. Leptoria Hoffmeister, p. 27. 1927. Leptoria Faustino, p. 141. 1928. Platygyra Matthai, p. 109. 1936a. Platygyra Wells, p. 124. 1936. Platygyra Yabe, Sugiyama and Eguchi, p. 38. 1939. Platygyra Umbgrove, p. 33. The genus having been known as Leptoria for 80 years, Matthai’s unravelling of Ehrenberg’s mistakes is nothing but a misfortune, a contravention of the rule that no change is to be made if greater confusion than uniformity will result. Matthai’s discussion of Ehrenberg’s confusion is a good example of the useless labour a rabbinical adherence to the rule of priority imposes upon men whose time is far too valuable for such rooting in old, and in this case, useless books. I therefore return to the older name, and bring my work into line with that of the eminent workers of the past. Matthai (1928), p. 112, writes : “ Platygyra further differs from Coeloria inasmuch as the corallum usually grows to a larger size, is heavy . . .” I find no record of the size to which Leptoria may grow except my own note of a specimen 2 or 3 feet across at Umm Qama’r, off Ghardaqa, which is much larger than numerous other specimens seen in that neighbourhood, on the outer reefs, and is trifling compared to numerous specimens of Coeloria lamellina. It is certainly heavy in comparison with C. lamellina , which, when dry, floats quite buoyantly. S. Manton’s largest specimen (pi. xv) is 45 x 36 cm. Leptoria phrygia (Ell. and Sol.). Four specimens are present, all labelled as Platygyra phrygia by Matthai. In some specimens the distinction from L. gracilis is very vague, as I found was the case in the Red Sea. The specimens are numbered 10, 328 June Reef, 42 and 443. Under the name Platygyra Stephenson and others give references to it on pp. 67, 86 MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 151 and 88, as one of the animals characteristic of the seaward slopes and anchorage, of Yonge Reef and the reef patch at Lizard Island. S. Manton refers to it on 5 pages and 5 plates. Colonies 3-5 inches across occur in the moat, but are more frequent and larger between 860 and 960 feet on Traverse I, but it is infrequent on the seaward slope of Traverse II. It does not penetrate into deeper water on the windward side (Graph, pi. viii). It is mentioned as occurring in the moat of Yonge Reef. The plates show only rare small specimens, except the one on pi. xv mentioned above ; it is evidently of no reef-building value here, or, apparently, elsewhere. Distribution : Northern Red Sea, Indian Ocean, including Mauritius, and Pacific as far E. as Samoa. Genus Hydnophora. Hydnophora microconos (Lamk.). One smooth little knob, No. 423, is all that represents the genus here, with a small specimen brought to me at Tahaiti by Mr. G. W. Otter, though there are a number of Australian records for H . exesa and II . contignatio. It is, however, possible that these species are not distinct. This specimen is matched by Matthai’s (1928), pi. xvii, fig. 1, for its general appearance and the monticles, but none of his enlarged figures show a similar thick and continuous columella, though fig. 7 on pi. xvi comes near. Vaughan’s (1918) pi. 47, figs. 3 and 3 a are similar. Local distribution as in the preceding species, appearing also as one of the massive corals of the outer moat of Yonge Reef. S. Manton (1935, pi. xiv), shows only two tiny colonies on the crest, and pi. xv three in the moat of Yonge Reef. Distribution : Northern Red Sea, throughout the Indian and Pacific Oceans as far E. as Samoa. [Family Merulinidae.] Genus Merulina. Stephenson and others mention the genus as one which, on the seaward slopes and the anchorage, are characteristic of vertical or overhanging surfaces. It was also found on the Reef Patch at Lizard Island. S. Manton refers to it on p. 289 as occasional in the moat, and shows a fan-shaped colony, 18 cm. across, on the seaward slope of Traverse 1 on pi. xi. Merulina ampliata (Ell. and Sol.). Two small specimens and four scraps are labelled by Matthai, and No. 349 is well matched by Matthai (1928) on pi. xiii, figs. 1 a, lb, and B.M. No. 296 by pi. 59, fig. 4 (Dana’s example of M. ampliata), except that the little knobby branches of the latter, which rise from the centre of the colony, are more marked in the Great Barrier Reef specimen, as are the deep radial wrinkles of the underside. In a number of specimens in the Kobenhavn Museum these curious knobby branches grow into large clusters, such as I have not seen illustrated, the nearest approach being Yabe, Sugiyama and Eguchi (1936), pi. 29, fig. 3. Thiel, pi. vii, figures a fairly large colony with numerous central branches, but of quite different form. I hope to discuss these Kobenhavn specimens in a later work. 152 GREAT BARRIER REEF EXPEDITION Specimen B.M. 296 is clearly Horst’s M. vcmghani, but Matthai (1928), pp. 130-131, pi. xiii, figs. 1-8, shows that the species grade into each other in respect to the sharpness or roundness of the ridges. In such a variable genus, and in all genera in which branches rise from a basal plate, only whole, or nearly whole, coralla are really worth examination. If no basal plate is accessible, as in some large specimens of this genus, or, e.g., Echinopora gemmacea, the col- lector should note the fact. G-ardiner notes huge colonies, such as no one else has seen, apparently without branches, “ off the outer slope at Goidu (Maldives) this or a similar species is very abundant, forming foliaceous colonies two or three yards across and high.” Distribution : Red Sea (in the North a small scrap, probably semi-fossil), Maldives, common in Malaysia, several records from the Great Barrier Reef, and as far E. as Samoa. [Family Fungiidae.] Fungia echinata (Pallas). No. B.M. 280 has no Expedition label or note of locality. It is part of a large specimen, apparently the half ; the piece is 15 cm. long x 13 cm. broad ; the whole would therefore be nearly as big as Doderlein’s largest from Singapore. Boschma in 1925 describes the difference between this species and Herpetolitha simplex Gard. In this specimen there is a tendency for the central groove to divide, it does not run out to the end of the calyx, all septa have well granulated sides, and the larger have well granulated teeth. The costal spines are conspicuously thick, of irregular length, spinulose, the costae themselves, though crowded, conspicuous from the edge to the centre, being clearly separated by deep slits, which, near the edge, make minute perforations. I find no figure exactly like this, Doderlein’s taf. x, fig. 2a, being nearest, but the spines are considerably smaller than in the Great Barrier Reef specimen. There are some large specimens in the Kobenhavn Museum from Singapore which are exactly like this of the Great Barrier Reef. Among them is one 39 cm. x 18-5 cm., apparently the largest recorded (cf. Doderlein, p. 102). Distribution : Red Sea to Tahaiti. Fungia actiniformis Q. and G. 1902. Doderlein, p. 82, taf. vi, figs. 1-10. The single specimen corresponds excellently with figs. 2 and 3 of the above plate. A photograph of an expanded specimen on the reef is given by Yonge in his book ‘A Year on the Great Barrier Reef ’ opposite p. 63, and in (1930) pi. i, fig. 4 ; and by Saville Kent (1893) in pis. 93 and 94. Distribution : As given by Doderlein only from the East Indian Islands, the Great Barrier, and as far east in the Pacific as Vanikoro (Santa Cruz Islands) and Rotuma. I distinctly remember it at Zanzibar, but it does not occur in the Red Sea or Tahaiti. Fungia scutaria Lamarck. Only one example of this easily recognized species was brought home, No. 275. The specimen is typical, with prominent tentacular lobes. Distribution : From the Red Sea to the Hawaiian Islands ; in Tahaiti along with the nearly related, if not identical, F. paumotensis (Boschma, 1929, p. 44). MADREPORAKIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA — CROSSLAND 153 Fungia paumotensis Stutchbury. Specimen No. 258 is labelled ? Fungia danai by Matthai, blit the equal-sized and equally spiny costae separate it from this species. It is round, 75 mm. in diameter, thick and heavy, with the septa 25 mm. high. The imder surface is flat but for depressed rings, 25, 13 and 7 mm. from the edge. The central scar projects slightly. I conclude that the specimen is abnormal, but nothing but its circular shape separates it from F. paumotensis. It corresponds especially well with Doderlein’s (1902), Taf. vii, figs. 1 and la. in the arrangement of septa and their conspicuous synaptacula, and the freely projecting ends of the septa. Distribution : Red Sea to Hawaii. Fungia fungites ( Li nn . ) . I place hi this very variable genus 5 specimens, Nos. B.M. 386, one not numbered, and Nos. G-.B.R. 86, 87, and P. 36. The first is labelled “ night Fungia,” the meaning of which I do not understand. They are between 75 and 145 mm. across. Of these G.B.R. 86 and 87 have been labelled £' F. repanda ” by Matthai and P. 36 F. fungites. Of the two former No. 86 is much like Doderlein’s figure of F. repanda as regards the upper surface, but it is the spines of the lower which are decisive, and these are (1) pointed, (2) smooth, except sometimes minute granules on their points only, (3) not aggregated into continuous fines and clusters as in Doderlein’s figures on pi. xii. I do not attempt to arrange these specimens under Doderlein’s varieties, as this has been found impossible by both Gardiner and Boschma, but remark that the smallest, B.M. 386, corresponds with Doderlein’s (1902) figs. 4, 4a, pi. xx ; the next, No. “ 0,” to figs. 7, 7a ; while P. 36, the only highly arched specimen, corresponds above with fig. 3, taf. xxi, but below is remarkably thickly covered with long pointed spines, not so crowded as to obscure the costae. No. 86 has the rather widely separated septa referred to above, the underside being like No. 0, but with larger spines, very like Doderlein’s (1902) fig. la, pi. xxii. No. 87 is similar but coarser. Distribution: I have already remarked in describing Forskaal’s collection (1941, p. 40), “ From numbers given by Gardiner (1908) this is much the commonest species in the Red Sea, and is distributed all over the Indo-Pacific, except the Hawaiian Islands.” Boschma (1929) records it from the Marquesas, but not from Tahaiti, the only Tahaitian record being Dana’s for his F. discus. It seems to diminish in numbers as it spreads east of the Great Barrier Reef. Fungia cyclolites Lamarck. 1902. Doderlein, p. 77 ; taf. iv, figs. 7-9 ; taf. v, figs. 5, 5a. 1925. Boschma, p. 205, pi. v, fig. 24 ; pi. vi, figs. 25-48. Stn. XIX, dredge, 26 specimens from 18 to 40 mm. x 34 mm. ; Stn. XXI, 3 specimens ; Stn. XXII, 10 specimens ; Stn. XXIII, 5 specimens. The smallest are flat, the older highly arched, as usual. The prominence of the principal septa varies, the specimens from Stn. XXII being flatter, and with principal septa less prominent than those from Stns. XIX and XXIII. Costae are generally scarcely visible to the naked eye centrally, but are always clear under a lens. Distribution : Red Sea, Indian Ocean, Gt. Barrier Reef, Philippines and China Sea, but not from Samoa and Tahaiti. 154 GREAT BARRIER REEF EXPEDITION Fungia patelliformis Boschma. (Plate XYI, fig. 1.) 1923. Boschma, p. 8, pi. ix, figs. 9, 11, 13-16a. Stn. XVII, 5 specimens (2 in alcohol), also 3 with a label now illegible. These correspond exactly with Boschma’s description and figures. Two are broken into segments, and one has regenerated from a segment only 13 mm. wide and 27 mm. in radius ; the regenerated circle being 43 mm. across, the original segment projects con- siderably from the regenerated part. I give a photo of the largest specimen, 52 mm. across, by transmitted fight, showing (1) the delicate yet imperforate base, (2) the junctions of the septa, (3) the columella. Distribution : Common in Malay region. Occurs also in Samoa and Hawaii, but not recorded from Tahaiti. Genus Stephanophyllia . This genus was placed in the Eupsamiidae by Milne Edwards and Haime (1848), p. 93, and by Milne Edwards (1860) p. 109, who remark that it is the only one with discoid shape and horizontal wall. In this they have been followed by all authors except Boschma, who describes (1923, p. 16) a new species found amongst Horst’s specimens of Fungia patella. I have never been able to see why Stephanophyllia should not be regarded as anything but a lightly built Fungia. Most species of Fungia have perforations, which, in Stephanophyllia are more numerous and wider than usual, a result of the fight building of many deep water forms. These clearly defined holes, bounded by stout beams, appear to me quite another thing to the general sponginess of the perforata, and made in quite another way. Synaptacula are also conspicuous. Stephanophyllia formosissima Moseley. 1881. Moseley, p. 201, pi. iv, fig. 11 ; pi. xiii, figs. 6, 7 ; pi. xvi, figs. 8, 9. 1907. Vaughan, p. 146, pi. xliv, figs. 2, 2 a. 1939. Gardiner and Waugh, p. 234. (Occurrence only.) One specimen from Stn. XV, dredge, B.M. 621, 25 mm. in diam. x 5 mm. high. It resembles Vaughan’s Hawaiian examples rather than those of Moseley in that the primary and secondary septa are as high as the rest, and the overlaps of the septal junctions do not extend far from the columella. The elongated columella is peculiar, in that, while obviously formed of septal processes, much of it is fused into a solid tuberculate mass pierced with small holes. , Distribution. — Kei and Philippine Islands, E. African coast near Pemba, Hawaii, now Great Barrier Reef. Herpolitha Umax (Esper). Only one specimen, from Lizard Island, “A” reef, No. 365. It is one of those in which the septa alongside the central row of mouths are specially wide and prominent, their edges making a pair of flat bands along each side of the groove. Horst (1921) gives, briefly, reasons for amalgamating the species Umax, foliosa, stricta and crassa, the survivors of the eight names given in his list of 31 references. Independently I came to the same conclusion in 1931 (Horst’s work not being accessible in Tahaiti), and MADREPORARIA, HYRROCORALLIXAE, HELIOPORA AND TUBIPORA — CROSSLAKD 155 gave results in rather more detail (1931. p. 354). I have a note that, in the Red Sea at Ghardaqa, all specimens differ from those of Tahaiti in the solidity and the regular toothing of their septa, and in that bending towards the centres is rare. S. Manton, p. 303, mentions it from Hedley and Taylor's work in 1907 on East Hope Isle ; Stephenson and others, pp. 67 and 88, list it as found on seaward slopes and anchorage of Low Isles as well as on the Reef Patch at Lizard Island. Distribution : Common a little below Low "Water Springs from the northern Red Sea to Tahaiti. Halomitra robusta Q. 1886. Podobacia robu-sta Quelcli. p. 140, pi. vi, figs. 5-56. 1898. Halomitra irregularis Gardiner, p. 528, pi. lxiii, figs. 1, 2. 1909. Doderleinia irregularis Gardiner, p. 282, pi. xxxix, figs. 25, 28. 1921. Doderleinia robusta, irregularis and sluiteri Horst, pp. 17-19, pi. iv, figs. 1, 2. 1925. Halomitra robusta Boschma, p. 242, pi. viii, figs. 99-104 ; pi. ix, figs. 107, 108, 112-16, 120, 122 ; pi. x, figs. 130-133. 1929. Halomitra robusta Boschma. p. 46. 1932. Halomitra robusta Thiel, p. 84 (with full synonymy), taf. xi, figs. 1, 2. 1939. Halomitra robusta Umbgrove, p. 45, pi. xiv, figs. 1, 2. Of five specimens labelled Doderleinia irregularis by Matthai, viz. “ P. 27,” and Nos. 19, 20 and 267, the fifth, No. 225, also labelled " portion of an attached funnel-shaped Fungiid is a fragment of Podobacia Crustacea. It has thick spinulose septal teeth and well roughened sides, a character of Doderleinia, which, as we have seen, is shared by several specimens of Podobacia Crustacea which we have examined. If Matthai means by this that all the genera Doderleinia, Halomitra and Podobacia should be joined, there is everything in favour of his view. The reasons for uniting all three species of Doderleinia under one, Halomitra robusta, are given by Boschma in 1925. I have also shown that the strength of toothing and spinu- lation of the septa is not even a specific character in Podobacia, and therefore cannot be a sufficient foundation for the generic separation of Doderleinia. Boschma’s account of variation in the genus leaves nothing to add . Of these specimens Nos. 19 and 267 are particularly rough and deformed, No. 20 is oval, somewhat flat, 23 cm. long x 13 cm. wide, with a distinct central mouth, the subsidiaries being radial and many concentric. Distribution. — East Indian Archipelago and Pacific, out to the Society Islands. Polyphyllia talpina (Lamarck). This species is not found in the collection but is recorded by S. Manton, p. 290, a large specimen 34 cm. x 7 cm., on the muddy sea floor at the end of Traverse I. The form is so distinctive that the identification may be taken as reliable. Distribution : Only from the East Indian Archipelago, Philippines and Murray Isl., Vanikoro, Santa Cruz Islands being the easternmost record. Genus Podobacia. The genera Halomitra and Podobacia have been united and again separated ; at the present time they have been placed in distinct families, following Vaughan (1905), siqrported by Gardiner in 1905 and 1909. 156 GREAT BARRIER REEF EXPEDITION We are on the horns of a dilemma : (1) The whole structure of species of the two sets is identical, down to small details. (2) Gardiner points out that, in spite of this, the union of the genera “ implies that the budding off of Halomitra from a fixed stock took place absolutely independently of that of Fungia, and that Podobacia is a form which has retained the primitive fixed character.” That is to say, it is impossible to separate the genera on any other character except that one lies loose while the other is attached, and impossible to unite them because of the extreme improbability of the evolution twice over of the reproduction of loose colonies from fixed stocks. The solution is that Podobacia is secon- darily fixed, the descendant of a free form. The method of reproduction by successive sheddings of coralla from fixed stocks was evolved as Fungia, from which Halomitra is readily derived, especially since Boschma showed how readily Fungia itself may acquire accessory mouths. Halomitra then gave rise to species in which the calices are retained on the stock, their multiplication spreading into a plate instead of the original free form adapted to lying on sand ; this fixed plate is Podobacia. The two genera are thus very closely related, if not identical. Some of the other fixed Fungiidae may derive from free forms in the same way, possibly Psammocora from this genus. Podobacia Crustacea (Pallas). (Plate XII, fig. 1 ; Plate XIII, fig. 3.) Horst (1921, p. 26) gives 14 references to this species, amongst which I find only the following two reliable figures (the last, by Bedot, is, as all his photographs, specially good, but it is probably a different species, though too young for safe identification) : (1) 1883, Duncan, Halomitra Crustacea, pi. vi. A young specimen with four rows of calices. Septo- costae arched. Underside with rounded ribs and well perforated. (2) 1905, Gardiner, p. 942, pi. 90, fig. 8. Gardiner had 20 specimens of this comparatively rare form from Goidu and Minikoi in the Maldives. It is not in the “ Sealark ” Collection. He gives only one figure, though stating that the variation in form, toothing of the larger septa and their granulation and spinulation, vary much. The figure shows septo-costae strongly arched between centres, which are near together. Details of toothing are missing. Faustino’s (1927, pi. 68, figs. 1 and 2) resemble this specimen, No. 36, more than do any other figures. He says that his example is like Dana’s, whose description is inadequate, and who gives no figure. Yabe, Sugiyama and Eguchi (1936, pi. 47, figs. 1-6) show excellently a specimen with thin arched septa, without teeth. Underside concentrically corrugated. The species has only one mention in the ecological papers, viz., S. Manton (1935, pi. xii) shows one specimen, 45 cm. across, near the base of the pinnacle. Specimen No. 36 is a portion of a saucer-shaped colony about 23 cm. across. The central calyx is not present. The difference from all but Faustino’s figure is in the flatness of the septo-costae, which begin at a low level on the distal side of each centre, rise gradually to a height of 2-3 mm., and drop suddenly to the next centre. The columella is a distinct style, rising from a flat fusion, in the well-developed centres. The underside of the plate is neglected in most descriptions ; in this case it is nearly smooth but for small spinulose projections, which sometimes fall into radial rows, but real ridges are found only within 1 cm. of the MADREPORARIA, HYDR OC'OR ALLINAE , HELIOPORA AND TUBIPORA— CROSSLAND 157 edge. In place of perforations are small scattered round holes, not arranged in rows between ribs, and not perforating the plate, unless within 1 cm. of the edge. As these peculiarities have not been described in other specimens I give illustrations on Plates XII-XIII. The figure of the upper surface is slightly oblique in order to show the shape of the septa and their toothing. As other photographs are taken perpendicularly to the surface they do not show details, so that Duncan's drawings remain the best illus- trations, but the cup-shaped ends of the !! ornamental granules of the free septal edge ” have not been seen again. Sample 225 B.M. 153 is a triangular fragment, 14 cm. measured circumferentially and 10 cm. radially, labelled “ portion of an attached funnel-shaped Fungid.” It is more normal than the other in that the septa are arched, project up to 5 mm. above the centres, are thick, have rough sides, and bear spinulose teeth. Columella none, or rudimentary, or bearhig 1 to 3 small points. No. 52 : The smallest, lightly oval, 15 cm. in diameter, no central polyp, edges bent over downwards, as some of Gardiner's specimens. Septo-costae flat, rarely arched, no teeth, only numerous spinulations along then edges, like those of their sides. Primary septa often thickened. Columella always rudimentary, at best has no style. Underside much as in Great Barrier Reef specimen. No. 46 : 20 cm., x 15 cm. A flat plate with two overlapping lobes. Base enwraps a dead branch of coral. Septo-costae sometimes arched, often flat, much thinner than in the Great Barrier Reef specimen. Sometimes toothed, but never so distinctly. In general much like No. 52. No. 43 : Two rather irregular plates from one base, like Faustino’s figure, 39 cm. x 27 cm. Edges folded. Much coarser with thicker septa than the other three. (This difference is therefore not due to age, as is already shown by Duncan’s figure of a very young one.) Septo-costae not toothed as a rule, but if toothed teeth are low and indistinct. Columella rudimentary. Underside as above, but in places large warts bearing the usual spinulose projections. No. 42 is a still larger specimen, folded in a complicated way. Some folds with very long septo-costae and very small centres, as hi the overfolded part of 52. Septo-costae thinner than in No. 43 or G.B.R. As No. 52 or No. 46, they project little so the whole colony is smoother. Teeth thin, often indistinct. Underside with high ridges, 1 to 2 mm. high and the same distance apart, bearing large teeth with spinules. There are from one to three much smaller ridges between the high ridges. All become fainter after 5 cm. or so, but in places can be traced to the base. Usually base covered only by small, denticulate monticles as in the other specimens. In one area these are enlarged, but do not form warts. I have also examined five specimens in the Museum of the University of Kobenhavn, all labelled P. Crustacea from “ Ostindien” or “ Singapuhra ” and dated 1858 to 1872. Specimen No. 36 is therefore a strongly marked variation, differing distinctly from the majority of those described and the specimens available to me for comparison ; but it seems unlikely that it is an independent variety. Distribution : Apparently always local ; only once recorded as abmidant (in the Maldives). Also Ceylon, Malaysia, Ryukyu Islands. Absent from the Red Sea and Pacific East of the Philippines. 158 GREAT BARRIER REEF EXPEDITION [Family Pectiniidae.] Genus Oxpora Kent. 1863. Tr achy for a Verrill, p. 53. 1871. Oxypora Kent, p. 283. 1879a. Echinophyllia Klunzinger, p. 69, pi. vi, fig. 8. 1905. Echinophyllia Gardiner, p. 949, pi. 93, fig. 26. 1935. Echinophyllia Crossland, p. 503, pi. 2. 1935. Oxyphyllia, Oxypora, Yabe and Eguclii, p. 429. 1936. Oxypora and Oxyphyllia Yabe, Sugiyama and Eguchi, pp. 50, 53. 1936.* Oxypora, Oxyphyllia and Echinophyllia Vaughan and Wells. [Check list of Generic names.] 1939. Oxyphyllia and Oxypora Umbgrove, pp. 40 and 41. The confusion in the literature of this genus has justification in its complication in nature. For an example of the latter see the figure (1935, pi. ii) of my Tahaitian specimen, but it is also due to the fact that, although all the colonies I have seen are large, all the illustrations are from fragments of the edges, just as the Great Barrier Beef collection consists of a number of fragments, the largest 8 cm. across. Consequently the five generic names, have been distributed among several families by Yabe, Sugiyama and Eguchi followed by Umbgrove. Vaughan and Wells, in their List of Generic Names of 15th June, 1936,* complicate the matter by recognizing both Oxypora Kent and Echinophyllia Klz. as valid, and Oxyphyllia as a synonym of the latter. This implies a difference between Echinophyllia and Oxypora which I am unable to find. I propose therefore to emend Verrill’ s definition slightly thus : Growth form in thin free plates, never encrusting for more than a small part of the base,! Calices on one surface only ; under-surface ribbed. Generally with slit-like perforations near edge and under centres, but these may be obscured by later growth. Septo-costae continuous from centre to centre, parallel between them, but at edges of centres bent towards the columella ; this is especially striking where a rib passes the end of the usually transversely elongated columella, and where they together form an H-shaped structure. Theca absent or imper- fect. Septa and septo-costae bear large spines with blunt ends ; these are structural, not due to incision. Columella trabecular. This definition may include other genera than those mentioned. It does not include Echinopora, which is an Astrean, whereas the septal arrangement of Oxypora is Fungiid. The generic distinction from such genera as Mycedium, Leptoseris, Podohacia, Halomitra, and others is largely in the spines of the septa and costae. Oxypora lacera (Verrill). Samples 28, 29 and 452 are all small scraps from young marginal growths. They show considerable variations, but there is no point in describing these without reference to whole colonies. I refer to the species under Verrill’s name and authorship, though the latter is an empty formality. The only real references are to Yabe, Sugiyama and Eguchi (1936), pi. xxxvii, figs. 1 and 2 ; and Umbgrove (1939), pi. xii, since no adequate written descrip- * Distributed privately in advance of the “ main body of the revision,” which was published in 1943. — [A.K.T.] t So far as present knowledge goes, and pace Ellis and Solander (1786), taf. 39. Compare Echinopora yemmacea, of which almost all examples are encrusting some with free edges. Exceptionally some may be free but for a small attachment. (See Crossland, 1935.) MADREPORARIA, HYDROCORALLESTAE, HELIOPORA AND TUBIPORA— CROSSLAND 159 tion of the species exists. It is unknown whether Klunzinger’s E. espera, Gardiner’s E. espera (sic) and Matthai' s Mycedium aspera are all the same species or not, and the material for determination apparently does not exist. A step might be taken if I had access to my Ghardaqa specimens. To foimd anything upon such accounts as are given by Ellis, Duncan, Terrill, Schweigger and Dana is impossible, Ellis’s figm’e is of the roughest, and his description nil. Terrill's description is good (with modifications) for the genus, but will include any species hereafter to be described. The first descriptions of any value are those by Klunzinger and Gardiner, but the identity of their species with those of Ellis and Terrill is apparently based on the supposition that only two species of the genus, doubtfully distinct, existed or would be discovered hi the future. Oxypora aspera (Ell. and Sol.). 1939. Oxyphyllia aspera Umbgrove, p. 40, pi. s, figs. 1, 2. I have pointed out the identity of these two genera, the latter of which Taughan and Wells identify with E chino phyllia. With the same proviso as in the preceding species I refer sample 84 to Umbgrove’s (1939) fig. 2 on pi. x, which it exactly resembles. Infor- mation as to what the rest of the colony was like is shown in Umbgrove’s fig. 1, though this is badly printed. The relationship to Mycedium is evident. One end of the small Great Barrier Reef sample No. 84 is thinner and more like the usual growth of the genus, and here are calices without walls and with long septal teeth. These teeth correspond to the upper thinner free portions of the septa conspicuous in Umbgrove’s fig. 2, as is shown by a single calyx of intermediate structure. Teeth on the costae are comparatively small in the sample No. 84, and in Umbgrove’s and Yabe, Sugi- yama and Eguchi’s figures, which makes it almost certainly not the species Ellis deals with. Genus Tridacopdiyllia. The name Tridacophyllia was exchanged with Pectinia Oken by Taughan in 1901, but as the former has been in constant use since 1830, the change can only cause more confusion than uniformity, and is therefore contrary to the rules of nomenclature. Matthai (1928), Yonge (1930), Thiel (1932), Yabe, Sugiyama and Eguchi (1936) and Umbgrove (1939), all retain the well-known name. Matthai (1928) p. 262 writes : “ This genus is usually placed in the family Astraeidae, but it seems to be related also to certain colonial Fungiids. Its systematic position can, therefore, be ascertained only after a comparative study of the latter group.” Yonge (1930), ‘ Physiology of Corals,’ I, p. 29, in this series, in a note says, “ 1 am uncertain whether this genus should be considered here or under the Mussidae ; in any case the two families have a great deal in common.” I unhesitatingly place the genus in the Fungiidae for two reasons : (1) The arrange- ment of the septo-costae and columellae of the secondary centres on the leafy expansions are certainly Fungiid, generally showing the characteristic H-form. Those of the main polyps are, as usual, regular, the genus differing from, e.g., Podobacia and Leptoseris in that there are many, instead of only one, primary polyp. (2) The method of feeding, as described by Yonge, is surely Fungiid. If it were of the Astrean or Mussid type, the mouths and minute tentacles, placed deep among the leafy expansions, would get no food, but, 160 GREAT BARRIER REEF EXPEDITION as things are, these expansions are themselves the food-capturing organs. The long series of specimens in the Kobenhavn Museum shows a curious morphological fact, viz., that the “ leaves ” are all enormously overgrown septa, so that of the hundreds of calices in a large colony, only a few primaries have any connection with the theca. This is quite clear in young colonies (such as those figured by Matthai in 1924), and the fully grown colonies may be compared to a bowl, an enomous theca, filled and overflowing with a mass of huge leafy septa, with many mouths, both principal and secondary. It of course makes no morphological difference that, in many cases, this bowl is cut up into branching segments. I hope later to describe this Kobenhavn series and give full proofs later. From this I am inclined to doubt whether such species as T. alcicornis Kent, T. cervicornis Moseley and T. primordialis Gard. really belong to this genus, since their outgrowths are thecal. The name T. alcicornis Kent has been given to a Japanese species by Yabe, Sugiyama and Eguchi, but I doubt its identity. A revision of the genus was undertaken by Thiel in 1932 (pp. 96-103), but it is rendered invalid by his taking the symphylloid form as primitive, and, apparently, deriving the genus from the meandroid astreans through this. His speculations on the biological value of the “ leaves ” were written before we had Yonge’s work on the feeding arrange- ments. The names given below are provisional. Tridacophyllia lactuca (Ell. and Sol.). 1786. Ellis and Solander, pi. 44. 1849. Milne Edwards, in : Cuvier, 1836-49, pi. 84. No. 417, a small distorted and partially diseased colony is probably of this species. The expansions bear secondary centres, and both these and the principals are loosely made by meetings of the widely separated septa, and neither have any columellae. All the expansions are thin. Distribution : Indian Ocean to Fiji, reaching its greatest abundance in Malaysia. Tridacophyllia paeonia Dana. 1927. Faustino, p. 160, pi. 42, fig. 1. (The only published figure giving details). No. B.M. 493 from Stn. XX (Yonge gives 6 fms. off Eagle Island) is labelled T. lactuca (Pall.) by Matthai. P. paeonia is probably only a variety, but provisionally I find that this specimen corresponds exactly with Dana’s and Faustino’s descriptions and Faustino’s figure pi. 42, fig. 1 . I add that the differences from T. lactuca appear to be (I) the thickness of the septa and their prolongations, which are as often pillar-like as plate-shaped, e.g., measuring 13x4 mm. in section, or round and 5 mm. thick and solid ; and (2) the well- developed columellae, of thin, closely packed trabeculae. The way in which the septa at the edges of the colony draw out the theca recalls Kent’s T. alcicornis. Distribution : Fiji and Southern Philippines ; now G.B.R. [Family Agariciidae.] Genus Pavona. The distribution in quantity of the species of this genus is exceptionally interesting. In the Northern Red Sea it is rare, the specimens seen at Ghardaqa numbering about four, MADREPORARIA. HVDROCORALLIXAE, HELIOPORA AND TUBIPORA — CROSSLAND 161 but it may grow to a considerable size, as examples in the raised reefs show. Similarly on tbe Great Barrier Reef, though Stephenson and others mention the genus five times (the specimens come from five scattered localities, viz., Low Isles, Three Isles, Lizard Island, and Yonge and Batt Reefs), S. Manton, in her detailed ecological surveys, met it only once (pi. xi). On the other hand, Gardiner found both solid and leafy forms common in Fiji and Rotuma, while in the Maldives only P. rarians ( P . repens Briig.) and other solid forms are important builders. In Hawaii only two solid forms are present, but not in any abimdance. Both forms are abundant in most places in the Pacific, perhaps reaching a climax in Tahaiti, where many lagoon reefs are practically made of leafy species of Parana and Parites ( Synaraea ) convexa alone. These leafy Pavonas are good builders ; they appear fragile, but in fact one can walk on most of them without doing much damage. It is deeply regretted that Matthai’s revision of the family was interrupted by the war. Parana decussata Dana. One specimen, No. 260, is the only leafy Parana in the collection. Hoffmeister (1925, pp. 40, 41) discusses at length the relationship of this species to P. danai, concluding that they are the same ; Vaughan (1918, p. 137) keeps them distinct, though remarking on their relationship ; his figure of P. danai on pi. 55 resembles P. decussata, and is widely different from the type of P. danai on pi. 56. Horst (1921 , p. 22) keeps the species distinct, though considering them “ nearly related.’' I leave the point without coming to a decision, only remarking that (1) Small scraps should all be destroyed, or at least, left out of con- sideration. (2) Vaughan’s figures of P. danai are objected to by Hoffmeister as being of very young specimens. Photographs of fully adult specimens of P. danai are published in my description of Forskaal’s collection, copies of which were sent to Vaughan and Matthai in August, 1939. The former agrees that Faustino’s P. cactus is the same as Milne Edwards and Haime’s P. danai, the latter that it falls within the limits of P. cactus. If so it seems as if Matthai were preparing to run nearly all the leafy species of Parona into one. We await his evidence, which we trust will be given in full. The present specimen is rather remarkable for the delicacy of the septa, the alternate ones being extremely thin as well as narrow, and for the prominence of the principal septa near the columellar pit, the latter feature giving the surface of the leaves a character visible to the naked eye. These are characters of P. renusta Dana (Vaughan, 1918, p. 136), a species never figured, until Horst (1922, pi. 31, figs. 1 , 2) did so. These show the narrow curled leaves described by Dana, so different from those in P. decussata. Distribution : Red Sea, Indian Ocean, Malaysia, Rotuma and Samoa. Parona danai M. E. and H. A single large leaf, 8 cm. high x 9 cm. broad, is thus labelled by Matthai. It differs from most specimens of P. danai in its thickness and weight, the small size and blunt edges of its marginal lobes, the few thick septa (the alternating thin ones being generally absent) and the absence of a columellar tubercle, the deeply placed columella being only an irregular fusion of septal ends. (See remarks under P. decussata.) Distribution : Northern Red Sea to Tahaiti, the latter locality based on an identifi- cation by Matthai. 162 GREAT BARRIER REEF EXPEDITION (?) Pavona cactus M. E. and H. Two small and very thick fragments : (1) from Batt Beef Patch No. 1 (the common species labelled by Matthai ? Pavona cactus (Forsk.). (2) No. 223 from T2 deep (without the ? mark). I agree that this is probably the species named by Matthai, but alter the author’s name to M. E. and H., since Forskaal’s species is P. danai. It would be correct to suppress the name cactus altogether, but this would cause much confusion. Distribution : Indo-Pacific. Pavona varians Verrill, 1864. (Plate XIII, figs. 1,2; Plate XIV, fig. 4.) Three very distinct forms of this astonishing species are present : Form A, No. 55 : The commoner of the three, but the most striking. It is an irregular mass, apparently a thick crust over branches of a dead coral. In places it is almost smooth, in others it has sharp ridges which, on the upper parts, are 5 mm. high (at one point 15 mm.). These bear calices after the fashion of the leafy species. I am surprised to find only one illustration comparable to this (Horst, 1922, pi. 31, fig. 3), as I found similar curious intermediates between plain crusts and leaves not uncom- mon in Tahaiti. Horst’s figure is not very good. In both this and Horst’s specimen the leafy expansions develop at the apices of lobes, as would be expected ; the high ridges result from vigorous growth in positions where horizontal expansion is no longer possible. We find the same thing in many other corals, e.g., Montipora, Merulina and Echinopora, and other encrusting organisms such as Polyzoa. Form B, No. 117 (Plate XIII, figs. 1,2): This is a rare form, so distinct that a special name may be desirable, in which case I suppose var. obtusata would be correct. Its synonymy has been confused by Quelch’s imperfect illustrations. 1886. Tichoseris obtusata Quelch, p. 114, pi. v, figs. 3-3c. 1898. Pavona calcifera Gardiner, p. 532, pi. xliv, fig. 4. 1905. Pavona repens Gardiner, p. 946, pi. xc, figs. 9-11. 1907. Pavona varians Vaughan, p. 135, pi. xxxviii, figs. 1, la. 1918. Pavona varians Vaughan, p. 138, pi. 57 (7 figs.). 1936. Polyastra obtusata Wells, p. 551, pi. ix, figs. 1. 2. The general appearance of this specimen is illustrated by Gardiner (1905, pi. xc, fig. 11) ; details by figures quoted below. It is uniform in structure over the whole mass, which is a rounded ridge 9 cm. high x 9 cm. broad, and about 6 cm. thick. Its infestation by a serpulid and several cirripedes has had no effect either on form or on details of the calices. The resuscitation by Wells of Ehrenberg’s and Quelch’s names, so deservedly sunk into oblivion, is nothing but a misfortune. The invalidity of the genus Polyastra had already been shown by data given by Gar- diner (1898), and a glance at Vaughan’s (1918) pi. 56, figs. 3 a, 36, would have been suffi- ciently convincing. Wells’s fig. 3, pi. ix of Quelch’s Tichoseris obtusata is so exactly like Gardiner’s 1898, P.Z.S., pi. xliv, fig. 4 of his P. calicifera as to return Quelch’s name to oblivion finally. This paper is apparently unknown to Wells. I should have been inclined to keep Gardiner’s species P. calicifera had he not himself in 1905 incorporated his closely allied P. intermedia into P. repens (which is P. varians ). MADREPORARIA, HYDR OC'OR ALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 163 No. B.M. 265 (Batt Reef, Patch. 1, Square 4) is similar but small, irregular and of coarser growth, and much infested by a cirripede. Form C : Sample 251 is a thin crust of a more usual form, in which the calices are in valleys, few having separate walls. Distribution : Xorthern Red Sea, Indian Ocean. Pacific to Tahaiti and Hawaii. [In the course of unpacking at the B.M. (Nat. Hist.) during January, 1950, of the Gardiner Bequest of corals from Cambridge there came to light what appears to be an important G.B.R.E. specimen “ 47.” On one side of the label in C. Crossland’s handwriting in ink appears : " Pavona ? varians, see other labels. G. "VY. Otter, Gt. Barrier Rf. Expedn.. Low Island.” In the corner is pencilled “ 47.” On the other side in G. Matthai’s handwriting in pencil is : Pavona varians (Verrill). This is an important specimen, as it shows the transition from the Pavona varians (Verrill) condition to the Pavona ponderosa (Gard.) condition. Note formation of corallites by intersection of radiating ridges and concentric collines. Photo — 47a, nat. size. Photo — 47 b, nat. size. G.M.” — [A.K.T.] Pavona duerdeni Vaughan. 1907. Vaughan, p. 135, pi. 38, figs. 2, 2a and 3. 1918. Vaughan, p. 135, table of species and footnote. 1922. Horst, p. 420 (synonymy), pi. 31, fig. 7. I consider that Horst is over-precipitate in combining this species with Dana’s P. clavus and Gardiner’s (1905) Siderastrea maldivensis. Vaughan, in his synopsis, gives a note that clavus, duerdeni and maldivensis are probably synonyms of P. exyplanulata, but no evidence has since been given. Sample No. 427 is a flattish stone, heavy for its size, rather irregular in outline owing to projecting angles (probably due to cirripedes and serpulids, of which a number are visible on the surface). It lay loose on the bottom, and is completely covered with healthy calices of practically uniform structure on every side and not modified by the presence of parasites. There are a few low rounded ridges in places, recalling P. varians. This specimen corresponds perfectly with Vaughan's of 1907 except in form. Septa of the third cycle, when present, are so thin as to resemble white hairs, and some of the ordinary septa have prolongations of the same hind. Particularly important is the structure of the walls ; all are thick and homogeneous, but while some are solid, others are plainly made of synapticula. In no case are they loose or thin as in Horst’s 2)1. 31, fig. 7, or Vaughan’s pi. 56, fig. 3 />.* The septa are never thick as in fig. 3a, nor are there projecting cushion-shaped sets of septa, or cut-off septa and costae as in Gardiner’s (1905) pi. 89, figs. 2 and 3. It is impossible to agree, in the absence of cogent evidence, that every solid Pavona with synapticular walls, whatever the form of those walls, of their septa, calices and costae, is of the species P. clavus. In fact I doubt if this species has been seen since Dana’s time. If the present specimen is a synonym of anything it is yet another variety of P. varians. Distribution : Hawaii and now G.B.R. Genus Pachyseris. The uniqueness of the skeleton needs no comment ; and for the complete difference between the polyps and their feeding methods from those of all other corals see Yonge, vol. i, p. 42. * Crossland’s typescript said “ fig. 7,” but he did not specify which paper of Vaughan he was quoting. The seventh figure on pi. 56 (1918) is marked 3 b, Pavona maldivensis (Gardiner). — [A.K.T.] 164 GREAT BARRIER REEF EXPEDITION Stephenson and others give references to it on pp. 67 and 87, as one of the corals characteristic of seaward slopes and anchorages, and of the fringing reef at Lizard Island. It is evidently rare, as usual, as S. Manton does not mention it. P achy sens speciosa (Dana). Sample No. 451, a deeply folded fragment, is typical of this species. No. 444, another fragment, has the same form of columella, hut the ridges are vertical, thin, high and sharp-edged. They measure (in sections) 5-6 mm. high, 2 mm. thick at the base, and 1 mm. at the top, which is rounded. There are 7 ridges in 2 cm. Only at the extreme edge of the plate are the ridges low, rounded and steeper on one side than the other, in the typical way. Two specimens from Singapore in the Kobenhavn Museum show the variation in the height and sharpness of the ridges noticed by other authors (Horst, especially, 1921, p. 35), but the highest hardly rival those of the Great Barrier Reef specimen. In one specimen all the ridges are high and valleys narrow. Distribution : Northern Red Sea, Indian Ocean and Pacific as far east as Tahaiti ; but not Hawaii. Pachyseris torresiana Vaughan. 1918. Vaughan, T. W., p. 132, pi. 55, figs. 1, la. Specimen 16, thus labelled by Matthai, is four small fragments, which fit together. The ridges are triangular, mostly very oblique, none high. Columella plate-like, inter- rupted by mouth centres. My only Tahaitian example of this genus seemed to me to combine the characters of this species with those of P. speciosa. The species is not recognized by Hoffmeister, while Horst thinks it may be identical with P. rugosa. I think the species should stand till more evidence is available. Distribution : Great Barrier Reef, Timor, and Tahaiti. Genus Coeloseris Vaughan. Coeloseris mayeri Vaughan. 1918. Vaughan, T. W., p. 139, ph 58, figs. 1-36. 1924. Matthai, p. 56 1936. Yabe, Sugiyama and Eguchi, p. 63 1939. Umbgrove, p. 53 Only one fragment, No. 405, 6 cm. across x 3 cm. thick, broken from a larger mass. It is evidently rare on the Great Barrier Reef, as it is not referred to in either of the ecological papers. This piece is all of the light type, with thin walls and thin septa, like Vaughan’s fig. la, so does not show the thickening of the inner part of the curved septa so often characteristic. Distribution : Bay of Bengal,* Batavia, Murray Island, Southern Philippines, Palau, Riu Kiu Islands and Taiwan. * The record of occurrence in the Bay of Bengal is due to Matthai who, in 1924, found a specimen in the Indian Museum labelled “ Arracan,” a place not to be found in an ordinarily good atlas. In an excep- tional atlas Arakan is found, a large estuary on the Burmese coast, 20° N. — a most unlikely place for corals 1 {These are records of occurrence only. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAXD ]65 [Family Siderastreidae.] Anomastrea irregularis Marenzeller. 1901. Marenzeller, p. 125, figs. 3, 3a. One tiny, but typical specimen, of 6 full-sized calices and some marginal buds on a mass of lithothamnion dredged from Station XVI. It differs from the type in the regularity of the calices, and in that the septa of the first cycle and part of the second are a little thicker and slightly more prominent than the others. These are the characters of youth. Distribution : Equatorial East Africa and Natal (collected by T. A. Stephenson). This is the third record. [Family Thamnastreidae.] Genus Psammocora. Psammocora contigua (Esper.). (Plate XV, figs. 4 and 5 ; Plate XVII, fig. 3.) Two small pieces, both alike, numbered 11 and 18, and a third, B.M. 417 “Low Island Reef” and No. P.25. The species is fairly common in the deeper part of the moat, but not elsewhere, but even in the moat it is common only in places (S. Manton, pp. 284 and 288, pis. iii and ix ; Stephenson and others, p. 49 ; under the name Ps. gonagra.) Three pieces are named Ps. gonagra Klz. by Matthai, a name rarely used since Klun- zinger’s time. Vaughan’s (1918) pi. 59, fig. 1, is difficult to reconcile with Klunzinger’s taf. ix, fig. 1 , with its thin septa, flat net-like coenenchyme, and much smaller calices ; but examination under a lens of the upper right-hand part of Klunzinger's photograph shows a greater resemblance. In fact the type specimen is abnormally variable ; having- been collected at Qoseir it is almost certainly a rough-water form, probably of Ps. contigua ; but discussion of synonymy without work on the reef is futile. It is curious that figures of this common and oft-described species should be incom- plete. Hoffmeister’s figures of Samoan specimens (1925, pi. 5, figs, la, 16) and Faustino’s (1927, pi. 70,) are the best, but the former is abnormally branched, and the latter shows only the outside of a colony. The present specimens 17 and 18 are so exactly like the one Esper figures that I give photographs corresponding to his figs. 1 and 2 (1795, pi. lxvi). These show the outer side of specimen 1 1 , and the inner side a broad flat branch with its distinct rows of “ stars ” ; also enlarged views of these latter. Those on the upper parts are similar, but rougher and looser. Distribution: Red Sea (?) ; through the Indian Ocean to Mauritius (Stephenson’s unpublished collection), Malaysia and the Pacific to Tahaiti and the Marquesas, but not to Hawaii. Psammocora exesa Dana. (Plate XVI, figs. 2 and 3 ; Plate XVII, fig. 4.) 1846. Dana, pi. 26, fig. 1 a-c. 1905. Gardiner, J. S., p. 952, pi. xcii, fig. 22. 1936. Yabe, Sugiyama and Eguchi, pi. xliv, figs. 3 and 4. One good specimen, No. 17, 14 cm. high and 10 cm. wide at the base, and a scrap, “ P. 26. Outside rampart Low Isl.” The latter is given this name with a query by vi, 3. 12 166 GREAT BARRIER REEF EXPEDITION Matthai. It is certainly this species, but is a younger growth, with looser structure, i.e., thinner septa, more spicular toothing and looser structures between the centres. Only in certain places does it resemble Gardiner’s figure quoted above, but neither in this figure nor the present specimens do I find “ interseptal spaces relatively large, at least of the same breadth as the septa.” This specimen might be worth describing in detail were it complete. It would not, however, form a link with Quelch’s “ Challenger ” specimen (p. 128, no figure), which I do not believe to be of this species. Specimen 17 corresponds well with Yabe, Sugiyama and Eguchi’s figures, which show part of a colony of columnar growth and calices (magnified 3 times). In the extreme coarseness of the septa, etc., it exceeds that of Gardiner’s and Dana’s figures. It does not, however, show the thick teeth, so closely placed as to leave only a series of narrow transverse cuts between them (pi. xvi, figs. 2, 3). The present specimen, No. 17, is nearly complete, as shown by its unattached spreading base (pi. xvii, fig. 4). It was growing- over two dead columns of Heliopora, and doubtless owes its form partly to this, but not entirely, as the three small humps on the basal expansion are quite independent. It is remarkable that the characters of the septa remain the same over the whole colony, except on the tops of the columns, where, as expected, the whole structure is looser and lighter, as in the other specimen. Distribution : Never abundant. Indian Ocean (Maldives) ; Pacific only as far east as Fiji. [Family Agathiphylliidae.] Genus Diploastrea Matthai, 1914, p. 72. 1816. Diploastrea heliopora (Lmk.), ii, p. 265. This species is one of the few corals immediately recognizable, and which has been given this specific name by almost all authors. The numerous figures published show no striking variation : only the toothing of the septa may be more or less pronounced. This uniformity may be correlated with the fact that the species was an important member of the Oligocene coral fauna, and it has therefore had time in which to become stabilized. It occurs in great masses where it establishes itself, but elsewhere it is rare. For instance, Umbgrove (1939, p. 42) never collected it himself in the Bay of Batavia, but saw it being brought ashore for road-making ! Similarly at Low Isles it is a rarity, mentioned once by Stephenson and others but not by S. Manton. The former authors (p. 88) refer to reef patches near Lizard Island, “ a notable mass of coral many feet deep, square yards of which were covered by a living colony of D. heliopora .” Distribution : From the Gulf of Aden (Jibuti) to Samoa ; but not reported from the Red Sea or Equatorial East Africa. It does not occur in Tahaiti. [Family Dendrophylliidae.] Balanophyllia incisa, sp. n. (Plate XV, figs. 1 and 2.) This species is branched, and therefore would usually be placed in the genus Dendro- phyllia, but, as Plate XV, fig. 2, shows, its branching is an entirely different thing from that of such species as D. micranthus, in which a special elongated polyp forms the stem MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 167 and its lateral buds the branches and twigs ; or such as D. axifuga , which is more or less dichotomous. I therefore retain this new species in the genus Balanophyllia. Two small broken scraps are present, which, by good fortune, fit together. There is no number but B.M. 369, and no note of habitat, but the following description of the colour is given on the label : “ Colony of a uniform combination of orange and carmine (nearly vermilion). Polyp carmine in its depth, modulating through fight shades to white at its edges. Base of colony brownish.” I have one complete, undamaged calyx, the half of one of its buds, and a much broken second bud on which to describe the species. Fortunately it is distinctive, the compression and lateral outline of the calyx being unique. The main calyx measures 17 mm. x 7 mm. and is 10 mm. deep if measured at the middle of one side ; but at one end the wall rises only 3 mm. above the columella and even less at the other. The half of the second, much smaller calyx, the first bud, shows a similar, but less pronounced form ; it is usual in most species for the younger calices to be more nearly round. The wall is highly perforated, the part above the columella translucent. Costal ridges corresponding to the 12 principal septa are prominent. Between them is a coarse branching reticulum with a mere tendency to vertical lines. On the buds the ridging is more regular. The lower part is nearly smooth, being covered by a dense epitheca. At the same time the walls, septa and columella thicken below, and, with the epitheca, make an almost solid stem. This, like the bases of the buds, is nearly round, and is 9 mm. in diameter. The 12 septa of the first two cycles are of nearly, but not quite, ecpial size. Between them are the 3rd and 4th cycles, the former short and cut off by the meeting of the septa of the 4th cycle at half to one-third the distance below the top of the wall. In one loculus a septum of the 5th cycle is present, in another both 5th and 6th on one side of the 3rd. The upper ends of the 4th cycle run close alongside and fuse to those of the first two cycles, and with them form the prominent projections of the walls shown on the plate. All septa are fairly thin and rough, with coarse tubercles, especially at their upper ends, which, like the walls, are abundantly perforated. Septa of lower cycles toothed, but these teeth appear to be perforations which have not closed up laterally. Balanophyllia yongei, sp. n. (Plate XIV, fig. 2 ; Plate XV, fig. 3.) Seven adult and four young specimens were obtained by dredging at Stn. IX. All grew alone but one, which seems to bear a lateral bud at its base, but some sort of abnormal epithecal deposit hides the actual junction. Two adult and three young are attached to lamellibranch shells ; one has settled upon a dead example of the same species (it is quite clearly not a bud), one is on a fragment of yellow calcareous sandstone (probably consolidated coral mud), and two have then bases broken off. Three of the specimens have been named B. bairdiana by Prof. Matthai. Milne Edwards and Haime’s (1860) description (III, p. 103) is too brief, and has no figure. We have only the account of Moseley (1881 , p. 190), who had access to the type, “ which is not in good condition.” The following are the differences from the present species. (Compare 1G8 GREAT BARRIER REEF EXPEDITION Plate XV, fig. 3, with “ Challenger,” pi. xii, fig. 4.) Ends of longer axis much lower than those of the short. “ Calicle in the adult irregularly elliptical in outline, being angular at the ends of the long axis.” The “ Challenger ” figure shows straight parallel sides with slightly pointed ends. The present specimens are generally regularly oval, but in three cases one end is rather blunter than the other. Principal septa slightly or considerably projecting, their costae decidedly the more prominent. They are straight, or sometimes slightly curved, and are twice as wide as those of the third series. “ Columella spongy, well developed, flat at the bottom of the fossa.” Milne Edwards and Haime say “ peu developpee.” It is fairly well developed and flat in these specimens, but less so than in Moseley’s figure (1881, fig. 5, pi. xii), and quite without the lateral extensions shown in this figure, and the diagram of the septal arrangements on p. 191. Even in those specimens with thick and very spinulose septa, where in consequence the fusion of the septa of the 3rd and 4th cycles is at its greatest thickness, “ their junctions are £ not ’ thickened by processes of the spongy columellar substance,” such as is shown clearly in Moseley’s plate. In normal specimens with thin septa the fusion remains thin and keeled, and the septum of the 3rd cycle runs distinctly through it to the columella ; on rare occasions it first joins, but does not interrupt, a 2nd cycle septum immediately above the columella. Milne Edwards and Haime’s type is about twice the size (area of calicle) of either Moseley’s or of these specimens. It is also significant that Moseley’s examples 'are from Bass Straits and Port Jackson, which are in a different current system to that of the tropical Pacific. Measurements of the 7 adult calices in mm. : No. 1. 13 x 10 x 14. 3rd order with entire margins. Septa thin, and finely spinulate. 2. 12 x 9 x 13. 3rd order toothed. Septa thin, and finely spinulate. 3. 12 x 9 x 16. 3rd order toothed. 4. 13 x 9 (broken). Sometimes toothed. Septa thin, and finely spinu- late. 5. 14 x 9x17. Slightly toothed. 6* 15 x 9 x 20. Always toothed. Septa thick, coarsely spinulate. 7. 13 x 8 (broken). Sometimes slightly. Septa thick, coarsely spinulate. * No. 6 is chosen as the holotype., — [A.K.T.] The heights cannot be measured accurately owing to the widely spreading base, which often includes projections of the substratum. These proportions agree fairly with Moseley’s measurements. On p. 191 he gives ratio of axes as 100 : 200, but the measurements of his larger specimen are 14 mm. x 8 mm., say 100 : 175. Milne Edwards and Haime give 22 mm. x 12 mm., or 100 : 185. Besides this variation in measurements there is a considerable other variation in these seven adult specimens. No. 3 is remarkable for the breadth of its principal septa, which, however, remain thin. In Nos. 5, 7 and 6, in this order, they thicken, become more coarsely spinulate and perforate, and spongy near the upper edge of the walls. The MADREPORARIA. H YRR OC'OR ALLIXAE , HELIOPORA AXD TUBIPORA — CROSSLAXD 169 projection of the walls at the junction of the principal septa is variable, usually very slight, but in No. 6 so considerable as to deserve illustration (Plate XIV, fig. 2). In this and the thickness and spinulation of the septa it approaches D. incisa. The younger specimens differ in the usual way from the adult. (1) The largest, 8 nun. x 7 nun., and 11 mm. high, is alone on a shell. Septa thin but with conspicuous spinules. Septa of second order joined by those of the third, and third by fourth high up the walls, the latter near the edge of the calyx. Columella well developed but loose. (2) A second, 7 mm. x 5 nun., growing on the side of No. 1, but probably not a bud. Septa thin and spinules small and inconspicuous. Septa of three cycles meet the columella, 3rd joined by 4th high up. Columella very loose. (3) The next smaller is damaged. It is practically round. (4) A roimd specimen, 4 nun. in diameter, alone on a small Anomia shell. Septa of second order met by the third high up on the wall. Fourth present as rudiments only in three loculi. Genus Dendrophyllia. The genus can only be defined as a colonial Balanophyllia, with colonies either encrusting or arborescent. Such a definition is unsatisfactory, being without morpho- logical foundation. See J. S. Gardiner in “ John Murray,” 1939, p. 237, on D. horsti, and compare the figures of this species on pi. ii with those of B. diffusa, immediately above them — a comparison which illustrates perfectly the difficulty of the distinction, especially as the septal arrangements of these two species are of the same type. It is conceivable that a distinction might be found between those species in which, in fully developed calices, the first three cycles of septa do not meet before joining the columella, and those in which some cycles meet, cutting off one or more of the lower cycles. Any two genera so founded would, of course, include both simple and compound species. The species fall into three groups according to colour : (1) Orange to vermilion, tentacles often pure yellow. Abundant as crusts and rounded masses under stones or corals in all warm seas — e.g., Cape Verde Islands — under bottoms of coal lighters and under stones just below Low Water Springs. At the mouth of the harbour in masses of branched forms at Low Water Springs, the only time it has ever been seen exposed to light. Panama, under stones, abundant, never exposed to light. Also in the Red Sea, in a similar habitat ; and generally through the Indo-Pacific, including Hawaii ; but none from the Great Barrier Reef collections or from the Murray Islands. Branched orange vermilion forms are from deeper water, and of these two species are here. (2) Black (the “ nigrescens ” group). In rather deeper water, not in the shade, branched. Stunted black specimens in habitat (1) are quite exceptional, if they ever occur. These species are very much more rare than those of group (1), and I have seen them only in the Northern Red Sea. (3) Bronm. In deep water, branched. Occasional brown specimens of group (1) are very exceptional. These colours are protective against actinic light ; the orange is exactly the colour of the paper in which photographic plates were wrapped 30 years ago. This light protection is obviously correlated in some way with the absence of zoochlorellae. 170 GREAT BARRIER REEF EXPEDITION Dendrophyllia arbuscula Horst. 1922a. Horst, p. 53, pi. viii, fig. 6. (Plate XIV, fig. 3.) Specimen B.M. 490, Stn. 9 and a number of small broken pieces (formerly in spirit which has dried up), which cannot be fitted together to make a second colony. I am attributing these fragments to this species in spite of differences. They agree in details of mode of growth, shape of calices, ribbing and the septal arrangement. (Compare Plate XIV, fig. 3, with Horst’s fig. 6.) They differ in being thicker in their stems, with the openings of the calices corres- pondingly larger, in the thinness and imperforation of the septa, on which lines of spinules are parallel to both horizontal and vertical edges. The columella may be a little smaller. A relationship to D. fistula (Alcock) is apparent. (Compare Plate XIV, fig. 3, with Alcock’s (1902, fig. 36, pi. v, of the type.) It may be that this species is merely a deep water — ooze habitat — form of D. arbuscula, differing mainly in the elongation of its calices and rarity of budding- — a kind of etiolation. A B C Text-fig. 1 .—Dendrophyllia arbuscula. A. Diagrammatic representation of the septal arrange- ment in a small lateral bud 4 mm. in diameter at the base of a long calyx. Four septa of cycle 2 are cut off by those of cycle 3, or at least do not reach the columella directly ; and two are complete to the columella. [From Crossland’s inking over of the septa of cycle 2 in his original pencil, ink and red chalk sketch it appears that the four septa of cycle 2 that he refers to as “ cut off by those of cycle 3 ” are those marked in this text-figure 1a by the number 2 in a circle, 2. A.K.T.] b and c. These diagrams represent adult calices whose longest diameter reaches 10 mm. In the great majority the arrangement of septa is as shown in b, but some- times it is as shown in c, where there are no fusions of the fourth cycle. D. fistula is described briefly by Alcock (1902, p. 42, pi. v) and by Gardiner and Waugh (1939, p. 237), and in detail by Marenzeller (1906a, pp. 8 and 16, pi. i). The three descriptions seem to differ fundamentally, as illustrated by the diagrams of Text-fig. 1, but the discrepancy is only apparent, as a study of younger calices of the Great Barrier Beef specimen shows. In these septa of cycle 3 may be found (1) running in to the columella without complication by those of the 4th cycle ; (2) joined by the 4th cycle, but proceeding to join the columella ; (3) apparently cut short by the 4th cycle ; septa meeting below it, as in the great majority of adult calices (Text-fig. 1b). In the youngest calices the 3rd cycle septa cut off those of the 2nd, but as Text-fig. 1a shows, the 2nd cycle septa may run right through the junction to the columella. In the great majority MADREPORARIA, HYDROCORALLIXAE, HELIOPORA ANT) TUBEPORA — CROSSLAXD 171 of adult calices the arrangement is as shown in Text-fig. lc, in which the 4th cycle appears to cut off the 3rd completely. Parts of a 5th cycle are present in some of the oldest calices, but there is nothing like the complication. Marenzeller figures for the older calices of D. fistula. There is no note of the colour of these specimens. Distribution : D. arbuscula has been found only once, by the “ Siboga ” in the Banda Sea at 45-90 metres. D. fistula, possibly a deep water form of this species, is found in the Red Sea, Indian Ocean, and in Malaysia. Dendrophyllia micrantkus (Ehrb.). Not Dendrophyllia nigrescens Dana. 1879. Coenopsammia micrantkus Klunzinger, p. 58, taf. vii, fig. 13 ; taf. x, fig. 13. I am compelled to note typical examples of this species, such as do not occur in the collection, before it is possible to describe the var. grandis which represents the species on the Barrier. Klunzinger gives the synonymy, and refers to Mihie Edwards and Haime’s (1857a) pi. E.2, fig. 2, of C. viridis, which, apart from its being “ green,” does not resemble this species very clearly. On p. 129 (1860) they give, under D. nigrescens, green, black and red species as synonyms, which I do not believe to be possible. They have only the barest mention (1848, p. 104) of Ehrenberg’s 0. micrantkus, merely quoting his two and a half fines of useless Latin description. Klunzinger writes “ nec Dendrophyllia nigrescens Dana,” in which he is right, though all subsequent writers are against him. I therefore can only refer to such later records as bear upon the distinction from Dana’s species, as, in the case of others, there is no knowing which species is referred to. The species is fairly common near Ghardaqa, in the Red Sea, but not as abundant as the orange and vermilion encrusting species (whether one or more), and never found in the same habitat, the black D. micrantkus being always exposed to the fight and never exposed at Low Water Springs : the orange species grows only in the shade under coral slabs on the outer reef edge, rarely on the inner reefs. The precise localities are the wall-like outer slope of the Abu Qalawa and Abu Fanadir reefs, and on Mortensen Rock, a great mass of decaying Porites in the middle of the lagoon. None of these Red Sea specimens, either Klunzinger’s, my own, or the five now before me belonging to the Kobenhavn Museum, exceeds 12 cm. in height. Klunzinger’s (1879) description and figures remain the only ones available. His figure (taf. vii, fig. 13) of the corallum is small, but being an actual photograph can be examined under a lens. The figure of the calyx (Taf. x, fig. 13) is not quite satisfactory. I wish to emphasize that features of the mode of growth which might be taken as accidental to Klunzinger’s specimen are, in fact, common to all that I have seen. Horst’s (1922a) figures are doubtful and too poorly printed to give any information. The only photographic figures of D. nigrescens Dana accessible to me are those of Vaughan (1918, pi. 60) and Faustino (1927, pi. 72), which show a quite different species to that of Klunzinger. All that Horst says (1922a, p. 50) is perfectly true, but the points he discusses are not those which really separate the species. The following differences are constant, and until further proofs and figures support Horst and Vaughan I regard the species as quite distinct. 172 GREAT BARRIER REEF EXPEDITION Colour : Dana (1849, pi. 27, fig. 1) gives a coloured picture of his I), nigrescens, differing from D. micranthus not only in form but also decidedly in colour. He shows a green black {micranthus is purple black, sometimes with a brownish tinge), tentacles of polyps white (fight brown in the text), disc bright green ; the polyps of D. micranthus are black uniformly. D. micranthus (Ehr.) : Bushy, few short stumpy branches. Height up to 15 cm. Branches flattened, or, if rounded basally, generally flattened at their ends. Outline of the principal polyp always more or less visible on the outside of the stem. Majority of the calices narrower at base and deep. Average breadth 7 mm. D. nigrescens Dana : Long straggling branches, often in one plane (according to Klunzinger, but see Horst). Branches round, 7-10 mm. thick. Majority of calices short and straight, comparatively shallow. 5 mm. in diameter. The appearance of ribbing outside the calices and down the stems, and character of the coenenchyme by which the stems are thickened, are variable in both species. The appearance of specially long calices in both may mean the beginning of a new branch. A group of them at end of a branch seems to indicate cessation of growth in length. As regards slenderness of branches and bushy shape, my Bed Sea specimens were all from sheltered water, and though some have more slender branches than have the speci- mens before me, none approach Vaughan’s figure, and are all built recognizably on the same plan as that of Klunzinger’s specimen- — viz., a main polyp forming a continuous tube from the base to the summit of the colony, which buds laterally, the larger buds originating side branches. As these buds often come off from opposite sides the branches become flattened, particularly at their ends, when growth in length slows down, so that nearly all colonies show the peculiar flattening of the ends of branches shown in Klunzinger’s fig. on pi. vii. Coenenchyme of a loose texture is added to the original calyx, so that the lower parts of branches are considerably thickened. As this addition tends to be more developed at the sides of the branches, particularly in the axils of buds and lateral branches, the round, main, original polyp-tube is more or less conspicuous in the flattened branches. This coenenchymal deposit may be ribbed longitudinally or be irregular (both conditions to the eye so different), often on adjacent areas of the same stem, and in some colonies the longitudinally ribbed form may cover the whole, from the youngest twigs to the base. The septal arrangements are, as described by Klunzinger, and by him alone, the simplest possible, viz., 6 septa of the 1st cycle rather broader than the 6 secondaries, and 12 (generally rudimentary) of the 3rd. The two 1st cycles reach a rudimentary columella of septal trabeculae, or sometimes only those of the 1st reach it, but in most calices, when fully grown, all the 12 septa of the first two cycles are equal in thickness and breadth and all reach the columella : sometimes even one or two of the thin 3rd cycle may also join the columella. There is no fusion of septa above the columella, but sometimes a septum of the 3rd cycle bends towards its neighbour of the second. In young calices the septa are thin and finely granulated on sides and edges, in older they are thick and coarsely granulated. In both they form a short triangular thickening at the wall : in the older they usually thicken decidedly just above the columella. These exceptions are to be recorded, but in the great majority of calices 12 septa reach the columella and 12 rudimentary septa alternate with them. MADREPORARIA, H YDROCOR ALLIXAE , HELIOPORA AMD TUBIPORA— CROSSLAND 173 Dendrophyllia miaranthus var. grandis var. nov. (Plate LY, fig. 1 ; Plate LVI, fig. 1.) This variety is represented by a single branch broken from a large colony. It is much regretted that more specimens, and at least one complete colony, were not brought home. The branch measures 15 cm. high, and side branches make the specimen 14 cm. wide. The broken stem is oval in section, 25 mm. x 17 mm. As shown on Plate LY the branches are nearly in one plane. The finer are, of course, the thickness of a single calyx, i.e., 7 mm., but the main stem retains its thickness nearly to the top. Details of the calices are as in the type species, but the columella is generally larger, and is of loose structure. In a few calices there are irregular traces of a 4tli cycle, and pairs of the 3rd cycle may actually join the 2nd between them. The most striking difference between this variety and the type, after the size of the colony, is the rough “ woolly ” surface of the steins, from a little below the terminal twigs to the base, though in most places a tendency to longitudinal ribbing can be made out. There is no note of the colour of the living polyps nor of the origin of the specimen, which is numbered 75, B.M. 500. Dcndroj)hyllia velata, sp. n.* (Plate LY, fig. 3.) A single branch of this minute form (habitat unknown) numbered only B.M. 390, but with a note on colouration : Colony orange, verging into brown towards its base. Polyps reddish brown.” In growth form it resembles Horst's I), minuscula (which does not agree very well with Bourne’s species, f especially as the latter has remarkably thick septa, and therefore cannot be young, as Horst says), but the columella and septa differ from both descriptions. There are also the curious hoods over the calices, recalling Stylophora, which are unique in this genus. The specimen is a fairly straight branch, 25 mm. long, of circular section, 4 mm. in diameter at the broken end, and that of the terminal calyx being 3 mm. This axillary polyp bears 8 buds, 6 of which are in pairs, but the pairs are none of them in the same plane, nor at right angles. Young calices are nearly sessile ; the youngest bud is near the lip of the terminal calyx, but the lowest on the stem are not the largest, and do not have the appearance of being the oldest. All are round, or very nearly so. The calyx walls are thin and per- forated, costal ribbing and perforations of the stem coarse for so small a species, and flaky. The characteristic hoods vary in tliire development, being absent from the terminal polyp and younger buds. In the uppermost pair one wall is higher than the other, rising gradually on one side, then dropping suddenly to the original level ; this makes the wall look as if broken, but examination shows that this is not the case. The lowest theca shows * Great caution should be observed in accepting this as a new species. Only one “ hood ” can be seen, and that appears to be a morbid growth, or perhaps the beginning of a new calicle. — [A.K.T.] t Crossland is evidently referring to D. minuscula Bourne, 1905, “ Report on the solitary corals collected by Professor Herdman, at Ceylon, 1902.” R. Soc. Rep. on Pearl Oyster Fisheries, IV, p. 213. — [A.K.T.] 174 GREAT BARRIER REEF EXPEDITION a similar development, but here the peak is at the side, not above the calyx when the branch is upright. The case of that of the longest calyx, one of the middle pair, is extra- ordinary, the hood being above and to one side, but also bending downwards so that much of the opening is hidden, and there is also a kind of crest above it. These curious arrange- ments can only be explained by figures. The septa are moderately thick, with fine spinules ; one to three of the primaries are especially prominent, secondary septa cut off from the columella by fusion of the tertiaries, this fusion often high above the columella, but sometimes near it, and sometimes not visible ; no traces of a 4th cycle, even in the cross-section of the main calyx tube given by the broken end of the branch. Here the septa are thickened so much as to be, in most cases, thicker than the spaces between them, but the addition to the outside of the wall is slight. The columella is remarkable for the height it rises above its junctions with the septa. It is relatively small, but well made of broad twisted ribbons with lightly scalloped edges. Genus Turbinaria. T. W. Vaughan (1918, p. 147) remarks : “ The species of Turbinaria reported from Australia by Bernard are listed on the following page. How many of the 27 reputed species should be recognized as valid can only be determined by a critical revision of Bernard’s original specimens.” Of these 27 species, 21 are from the Great Barrier Beef and Torres Straits, 4 are from West Australia only, and 6 are common to E. and W. The genus is mentioned by Stephenson and others (p. 67) as characteristic of the seaward slope and anchorage, on vertical or overhanging surfaces below the level of low water, and both massive and foliose species on the reef patch at Lizard Island. S. Manton refers to it four times, and figures a single specimen at the base of the pinnacle of pi. xii. As her diagrammatic sign is a cup with a thick border the species is almost certainly T. peltata. It is among the first corals to appear on the seaward slope of Traverse I, and minute specimens are among the corals which survive the mud at the foot of the slope and on the sea floor. It appears again, but rarely, near Traverse II. Apparently the genus was found to be more common by Hedley and Taylor on the lee side of East Hope Island . Distribution : The genus is not recorded from Samoa,* Tahaiti, or Hawaii, though common from the Bed Sea to the Western Pacific. Gardiner (1898, p. 262) notes : “ There is a marked absence of this genus both at Funafuti and Botuma, only one colony having been found, while in Fiji three species were obtained.” Turbinaria peltata Esper. (Plate XVI, fig. 5 ; Plate XVIII, fig. 2 ; Plate XIX, fig. 2.) 1896. Bernard, p. 38, pis. vi, vii, viii and xxxi, fig. 15. Sample 435 (major diameter 18 cm., height 8 cm.) is a stage between those figured by Bernard on pis. vi and vii, in which the edge of the cup is folded (one fold fused to a two-sided plate) and there is one upgrowth from near the centre. Having a second variety with which this is to be compared it is necessary to fill in some details of Bernard’s description. Except one doubtful record, T. frondens, by Bernard (1896, p. 46) ; see note on p. 93. MADREPORARIA, HYDRQCORALLTNAE, HELIOPORA AND TUBIPORA— CROSSLAND 175 The aperture is circular except when the calices are crowded. The majority are 4 to 5 mm. across, or 6 mm. by 4 mm. if oval, the thin walls adding 1 nun. (as in the three examples figured by Bernard and the one by Faustino). Milne Edwards and Haime’s 10 mm. must refer to a giant calyx, such as is not present in this or in Bernard's collection. According to Bernard the septa are 12 primary and 12 secondary. In this specimen I distinguish 6 primary, 6 secondary, and 12 tertiary, the total number of 24 being constant. At first sight all seem to be of the same width, as in most species, but more careful examination shows that the 6 primaries are a little broader than the secondaries, and these than the tertiaries ; minute points and ridges indicate a rudimentary and incomplete 4th cycle. The septa are often toothed on then upper quarters, and are slightly granular on their sides. The septa being narrow, the columella is large, and, even in round calices, generally oval, the septa at each end slightly narrowed. It is composed of broad twisted leaves, thin and well separated, on the whole extending upwards to form an irregular honeycomb, but in many cases a central leaf forms a ridge along the top of the columella, which may become conspicuous and to which the lateral leaves may be joined pinnately. As published figures show this structure imperfectly or not at all, I illustrate it on Plate XVIII, fig. 2. The coenenchyme is spongy, composed of broad, thin, twisted plates, more or less horizontal, bearing a few small, blunt upward points ; in rare places these run into irregular sinuous lines, on which the teeth are more numerous, with grooves between — “ a well marked gyrating system of ridges and furrows,” as Bernard describes it. On the underside this arrangement is the rule, the ridges running radially, and the upright processes are more marked, though still low and blunt. There are numerous round or oval holes, resembling worm-holes. Saville Kent (1893, p. 188) describes the colour of the living colony as whity-brown, whitish polyps with greenish centres ; tentacles numerous and simply subulate ; both occasionally a delicate rose pink. Dana (1849, pi. xxx, fig. 4a) shows the colours as brown, yellow or greenish brown with bluish discs. The colours of other Turbinarians, though often brilliant, are variable within limits (see Waugh on T. mesenterina, 1936, p. 913). Distribution : Bernard (1896, p. 38) says : “ This species is the most striking in the genus Turbinaria, and also the commonest.” This latter is due to the diligent collecting by Saville Kent. Nearly all Bernard’s localities are from the Great Barrier Reef and Torres Straits, but some are also from W. Australia and Singapore. Most other records are from Malaysia, but Dana, Milne Edwards and Haime and Faustino took it as far east as Fiji and the Philippines. The only Indian Ocean record is Mauritius, probably as unfounded as that from the Red Sea referred to in a note by Bernard. To the evidence for its absence from that sea I add Gravier’s, Waugh’s and my own. I have here from Mauritius a very different species, T. plicata, collected by Stephenson. Turbinaria peltata var. gibiari n. var. (Plate XIV, fig. 1 ; Plate XVI, fig. 4 ; Plate XVII, fig. 1 ; Plate XVIII, fig. 1 ; Plate XXII, fig. 2). Sample 193 is a sector from the edge of an explanate or cup-shaped colony measuring 11 cm. along the base and 6 cm. radially. The broken base is 17 mm. thick in the middle, 7 mm. close to, and 4 mm. at the extreme edge ; it is gently folded. 176 GREAT BARRIER REEF EXPEDITION The differences from the normal are : (1) Calices projecting obliquely, those near the base 10 mm. on proximal side, near edge 5 mm. ; the distal sides 3 mm. and 1 mm. (2) Columella as in preceding, but the directive ridge may rise high above the rest. The coenenchyme is covered with thin plate-like ridges bearing numerous minute but distinct pointed teeth, which, under the binocular, are seen to be set transversely to the plate. The underside is grooved, the grooves long or short, more or less sinuous. On their walls are similar plates, but thicker, shorter, and with less distinct teeth, often two rows of plates on each wall. Near the edge the round holes are deep and numerous, centrally rare and shallow. Specimens with obliquely projecting calices are not mentioned in the literature, but there are three large examples in the Kobenhavn Museum, labelled T. lactuca n. sp. by Liitken. They differ from the above in that the columella never has a directive ridge, the 1st and 2nd cycles of septa are hardly distinguishable and in the coenenchyme. I hope to describe these and other specimens of the Kobenhavn Museum later. Possibly Liitken’s unpublished name will stand, if only as a variety. Turbinaria bifrons Briigg. (Plate XXI, figs. 1, 2.) 1896. Bernard, p. 69, pi. xxi, pi. xxxiii, fig. 1. Specimen B.M. 272 is whole, 13 x 20 cm. and 16 cm. high, and is evidently one of those referred to by Stephenson and others as growing on a vertical surface. Specimen B.M. 413 (Stn. XXIII dredge) is a single “ leaf” 10 x 10 cm., with two fragments. The species is easily recognized by the complete fusion of its folds, its spiny septa and fused (“ glassy,” Bernard) columella. The obliquely placed thecae are also a constant feature. Specimen 272, in its complicated folding, shown on pi. xxi, resembles the “ two minute stalked specimens ” of Bernard’s pi. xxi, rather than the larger one of the same plate. A curious feature is the occasional formation of small round branches, which superficially resemble an Acropora by the oblique calices. The doubling of the plates is not all due to folding, as fig. 1 shows the outside of the colony is also overgrown by calyx-covered leaves, though not always completely. In this complete specimen the spinulation of the sides of the septa is usually less marked than in Bernard’s pi. xxxi, fig. 1 ; but in the dredged fragments the septa are both thicker and more spinulate. The columella is not always prominent, and often appears folded or granular on the top. Distribution : West Australia (young examples, Bernard). Also, if synonymous with T. cons])icua, already recorded from the Great Barrier Beef. Turbinaria frondens (Dana). (Plate XXIII, fig. 1.) 1846. Gemmipora frondens Dana, p. 412 ; 1849, pi. 27, fig. 10. 1860. T urbinaria frindescens Milne Edwards and Haime, p. 167. 1896. Turbinaria frondens Bernard, p. 46 (no figure). MADREPORARIA. HYDROCORALLIN.AE, HELIOPORA AND TUBIPORA— CROSSLAND 177 Possibly T . brassica (Dana) and T. danae Bernard are the same species, but the latter seems to be distinguishable by its broad septa, the former by its shallow calices and smooth coenenchvme : a series of specimens might possibly bring them together. The two pieces here, Nos. 326 and 383, both from Lizard Island, Beef A, agree completely in these respects, and differ from a specimen in the Kobenhavn Museum, which constantly agrees with T. danae. Dana’s description being altogether too brief, and his figures often inaccessible, I redescribe the species.* The specimens before me are two plates, both folded and one deeply wrinkled at one side, measuring 12 and 17 cm. radially and 16 and 10 cm. across. I take the obliquity of the calices to indicate that they grew more or less vertically. One of them has a small branch 15 mm. high, growing from the surface of the plate after the style of Bernard's T. aurantiaea ; its seven calices are all on the distal, or upward, side. Another projection surrounds what looks like a mollusc burrow. The obliquity of the calices is shown in figure on Plate XXIII ; even the immersed calices of the proximal part are also oblique at their openings. Near the edge the thinness of the plates allows the lower parts of the calicinal tubes to form ridges, which are also visible on the outer side of the plate (which is thus wrinkled). Lower down the coenen- chyme is rather thicker, and the calices, cylindrical above, become elongated cones : close to the base they are, as usual, immersed. They are slightly oval, 2 mm. in larger diameter ; the walls are thin. The septa are always narrow, though a little broader in the proximal calices, at first sloping steeply, then dropping vertically, leaving a wide open fossa, of some depth. They are practically all alike, 20 to 24 in number (most often 22), thin and smooth. The columella is broad and oval, only slightly arched above. It is formed of com- paratively large foliae, twisted in the usual manner, which do not make a directive ridge. The coenenchyme is covered with short ridges, bearing granules, or blunt and indistinct teeth, often set transversely. The ridges are thicker than those shown in Dana’s fig. 10c, but have squarish knots in them such as he shows. They are thinner and more regular on the back of the plate, which, to the naked eye, appears smooth. There are no holes. Distribution : Dana’s specimens were from Fiji. Bernard records one from Samoa (presented by Bev. S. J. Whitmee), but no Turbinarian was found there by Mayor, and missionary records must be received with caution as, in past days, they carried on a trade in corals as curios. Turbinaria sp. Specimen No. 7 is a small flattish cup, 9 cm. in diameter, with a very eccentric stem ; this eccentricity, together with the appearance of the calices towards the edge of the higher rim, suggests that it is the young form of such a growth as T. frondens. It is not of this species however. I consider it rash to name so young an example. Matthai suggests (with a ? mark) T. aurantiaea , but it differs from this species thus : * I owe my copy of Dana’s work to the Librarian of the Zoological Museum of the University of Kobenhaven, who obtained the loan from the Royal Academy of Science, Stockholm. I am deeply gratified by this example of international co-operation. 178 GREAT BARRIER REEF EXPEDITION T. aurantiaca (Bernard, p. 33). Calices 1-5 mm. or less . Not typically projecting Septa 18-20 . Columella spongy, compact This specimen. Regularly 2-0 mm. Most project 1 mm., but openings are not oblique. Septa 20-24, usually 20. Distinctly spinulose sides. Columella formed of thick, close-fitting, granular leaves, usually transverse to oval. No directive ridge. The septa are not like those of T. pocilliformis. In calicinal characters it comes nearest to T. auricularis, and there is no reason why this should not be the start of a large growth like those of this species. I am confirmed in my refusal to name the specimen by comparison with one of the same size in the Kobenhavn Museum, and almost identical to the naked eye, but the calicinal characters of which are completely different. Glomerate Turbinarians. Although I cannot follow Bernard in his emphasis upon growth form in this genus these solid glomerate forms are probably separated by a real morphological difference in the continued growth of their polyps into long tubes. Bernard himself writes (p. 166), “ I was at first disposed to look upon these as merely massive varieties of other types,” but in the only species of which the growth variations have been specially studied, T. mesenterina (Waugh, l.c.), such elongation of the polyps does not occur. Pending further investigation on the reefs, I therefore regard this as a distinct division of the genus, and the species about to be described as characterized by, among other things, their mode of growth. To describe two new species from single specimens is most uncongenial, but I have no choice. Turbinaria stephensoni, sp. n. (Plate LV, fig. 2 ; Plate LVI, figs. 2, 3.) This might be a variety of T. stellulata Bernard,* but there are constant characters of the septa and columellae not mentioned by Bernard, and which are unique in the genus. Specimen 449 is a rounded hump, 75 mm. in diameter x 50 mm. in height, growing on a piece of rotten coral, probably a solid Montipora. By prising this away I find that the hump, though concave beneath, and spreading downwards into an encrusting sheet, is, in reality, 30 mm. thick. There is also a thinner extension on one side of the base about 3 cm. wide and 10 mm. thick (near the edge about 5 mm.). The calices project slightly and are conical, conspicuously so on the top of the hump, where they may project 3 mm. and be 3 mm. across on the top, though the openings are only 2 mm. On the lower expansions they project little though remaining conical, and the opening is generally 1-5 mm. across. It is slightly oval. The calyx wall is well defined. The coenenchyme bears little plates and ridges, with sharp-pointed processes in little groups, a distinct contrast to the blunt granules of the next species. The number of septa averages 25 on the hump, 20 on the lower expansions. They are thin, very finely toothed along their edges, their sides very finely spinulate, thicker * Bernard, p. 65, refers the species to Blainville and Lamarck, but on very slender grounds. The only way to define the species is by using Bernard’s name. JLADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA — CROSSLAND 179 and more spinnlose on the spreading bases. They are very narrow above, and slope down to the columella either without forming an angle, or forming at most a low and rounded one. The septa share with those of T. stelhdata a peculiarity in this genus, that they are not all the same length : short and long do not regularly alternate, the short being much the fewer. A further peculiarity, not mentioned by Bernard, is that septa adjacent to a short one are often bent towards it, and sometimes (generally once in each theca) fuse between the end of the short septum and the columella. In distinction from T. stellulata most of the long septa very clearly join the columella, and sometimes bear a knob just before they join. The columella is also very distinctive. At its best development it is small, oval, not arched above, elongated in the same direction as the calyx, solid and tuberculated above. A hue of the small but high tubercles forms a crest along the top, which often developes into a high plate. In many cases the lower part is inconspicuous, often invisible, the plate then remaining the chief, or even the sole, representative of the columella. Ordinarily this is quite distinct as a row of tubercles, but, towards the edge of the colony it may become irregular and fused ; in a few calices it degenerates to a tubercle or two, or may be absent, the septa then meeting in the centre. Where the columella is less developed the lower teeth of the septa are more prominent, as would be expected. Turbinaria mantonae, sp. n. (Plate XVI, fig. 6 ; Plate XVII, fig. 2 ; Plate XIX, fig. 1 .) Specimens 323 and 316, both labelled “ A ” reef, Lizard Island. To the naked eye the specimens are distinguished from the preceding by the slightly projecting cylindrical thin-walled calices, sometimes level with the surface they are 2-5 mm. across the openings, with many smaller, down to 1-5 mm.* and even 1-0 mm. (fully developed) on the sides low down ; on the hump are also numerous small calices, but these are young. They are distinctly deep. One hump of No. 323 is broken through and shows thecal tubes 5 cm. long. The specific gravity is high, perhaps more so than in T. Stephens mi. The coenenchyme is a fine but loose network on the summit, but in places gives place to more regular, wavy lines bearing low blunt granules. Fully developed calices are completely regular, oval or round, of a type common in the genus. The septa are all alike, sloping gradually at first then dropping vertically to the columella. They are thin and smooth, and in full-sized calices their number averages from 24 up to 27. The columellae are large, oval, of closely packed trabeculae, lying deep in the fossa, flat on the top and bearing numerous upright tubercles. In many calices the lower parts of the septa are toothed and bend suddenly towards the centre, passing into the columella, on the top of which they form low ridges. In younger calices the septa remain distinct almost to the centre of the columella. Thus we have here two cases in which the columellae are seen to be made from the septa, a fact which seems to be not usually visible in this genus. * Corresponds in these respects with T. parvistella Kent, which cannot be identified from the descrip- tion given. The name therefore lapses. ISO GREAT BARRIER REEF EXPEDITION [PERFORATA.] [Family Acroporidae.] Genus Astraeopora. As an example of the difference between pure museum work and observations on the reef compare Bernard’s division of this genus into explanate, pulvinate and glomerate forms with Waugh’s demonstration that all these forms frequently occur in one colony. His glomerate division is particularly artificial. Such detached forms occur in almost any sort of coral, and is purely environmental. This is especially well shown by the numerous potato-like loose growths of Leptastrea, Cyphastrea and Porites, which abound in a certain habitat near Ghardaqa — viz., the wide, weedy flat on the west side of Abu Qalawa Reef. Bernard’s Australian species are listed by Vaughan (1918, p. 145), who accepts four of them, describing A. myriophthalma (Lmk.) and A. ocellata Bernard, while A. profunda Verrill is redescribed, after examination of the type, by Hoffmeister. When in Tahaiti I tabulated these species for comparison with the single good-sized specimen I found there. There may or may not be a columella (the distinguishing mark of A. profunda according to Hoffmeister) and the primary septa may or may not meet, and, in short, this specimen seems to combine the characteristics of all these three species. Indeed, of Bernard’s 14 species it seems unlikely that more than three or four are real. The genus is listed by Stephenson and others on pp. 67 and 86 as characteristic of the seaward slope and anchorage at Low Isles and of Yonge Reef. By S. Manton it was found on the muddy sea floor (two quite fair-sized colonies 16-20 inches across) and on the slope where the influence of the mud begins. Astraeopora myriophthalma (Lmk.). 1930. Astraeopora ocellata Yonge, p. 46. 1936. Astraeopora myriophthalma Waugh, p. 927. Specimens 41 and 329, the latter from Ribbon Reef, are typical of massive growths of this species. No. B.M. 392 and No. 428 are of different appearance, but, in view of what I have written above I consider them the same. No final revision of the species is possible without long investigations on the reef, and an extension of Miss Waugh’s work to other seas. Both Yonge and Waugh give figures of the living polyp which differ widely. It seems more likely that Yonge’s specimens were not fully expanded than that the species A. ocellata can be distinguished by having short stumpy tentacles. At the same time the naked eye anatomy of polyps may very well be of use in the descrimination of species. Preliminary observations at Ghardaqa showed that this is possible for some Faviidae. Distribution : Though not common in collections the species has been recorded from the whole Indo-Pacific, from the Red Sea to Samoa and Tahaiti. Genus Montipora. The usual division of this genus, introduced by Bernard, into smooth (glabrous), foveolate, papillate and tuberculate is generally useful, and the only one possible ; but the 19 species here described include some which are both smooth and foveolate — e.g., M. ramosa ; and tuberculate and papillate — e.g., M. prolifera. In foveolate species the MADREPORARTA. HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAXD 181 ramparts may occasionally break up into papillae, and Bernard's class “ Papillae irregular ” includes suck forms as M. venosa, which are often more foveolate than papillate. It is just possible that papillae formed from ramparts are not homologous with those formed independently, but this is hardly a practical distinction. I have not attempted to trace out all possible synonyms, contenting myself with the few which are well grounded and illustrated by the present collection. Doubtful and possible synonyms are not helpful, witness the wild confusion caused by baseless guesses at the identities of ForskaaFs Madrepora monasteriata and M. rus. — the trouble by no means ended by Marenzeller’s showing what these species really are. As in the case of Acropora and the other perforates most of the species are confined to the Malay-Fiji area, in contrast to the imperforates, many of which are found over the whole Indo-Pacific, from the Bed Sea to the Tuamotus, dying out in the Marquesas. Smooth, explanate Montiporae. Montipora granulosa Bernard. (Plate XXV, figs. 1,4: Plate XXVII, fig. 4.) 1897. Bernard, p. 21 ; pi. i, fig. 2 ; pi. xxxi, fig. 3. Bernard described 14 glabrous, explanate species : other authors very few. The present specimens do not completely correspond with any of them, but come so near to M. granulosa that I give them this name and detail the differences from Bernard’s single specimen from 44 fms. on Macclesfield Bank. These, No. G.B.B. 113, dredged from Stn. XXV,* consist of 4 small irregular thin plates, the larger 8 cm. across its base, 9 cm. radially ; they may or may not have formed one corallum originally. Unlike Bernard’s specimen they grew freely, no point of attachment is present, the only approach to one being portions of older lamellae of the same species. The epitheca is well developed, generally right up to the edge, which, in some places, bends back and forms a thin, living layer over the epitheca for from 5 to 10 mm. The epitheca is strongly marked with fine and coarse concentric wrinkles, besides deep wrinkles in the corallum itself. These were caused by periods of cessation of growth, as shown by small portions of the free edges projecting from the epitheca at some of these folds. The surface is raised into low rounded hillocks, often elongated radially or circum- ferentially, and easily distinguished from those resulting from adhering Balanids and Serpulids. They do not bear calices, all of which are level with the more or less flat surfaces between these hillocks, nor is there any relation suggesting modified foveolae. Calices correspond with Bernard’s description, except that none “ appear to open irregularly on slight mounds of coarse reticulum.” Some are irregularly compressed. The “ granulations ” of the surface extend onto the proximal parts of the septa. These “ granular ends of reticular threads ” deserve more description than Bernard gives ; as he says, they “ give a soft velvety sheen to the whole surface,” which is characteristic of these specimens, and of no others in the collection. With few exceptions — e.g., close to the growing edge — they cover the whole surface thickly ; they are in the form of minute plates bearing a row of 2 to 5 vertical slender spinules, or else a little clump of the same ; between them can be seen the rounded pores of the reticulum. * Comparatively shallow water near the Low Isles may be ecologically equivalent to much greater depths in the open sea — e.g., at Macclesfield Bank, vi, 3. 13 182 GREAT BARRIER REEF EXPEDITION The section shows a thin glassy solid layer, not distinguishable from the epitheca (a narrow reticular layer), above which most of the section is occupied by thin irregular vertical rods joined by horizontal nodes. Bernard describes them as short. This is, of course, the case where the lamina is thin ; near the growing edge the reticulate layer occupies most of the section. Distribution: One specimen from Macclesfield Bank in 1888. Now Great Barrier Beef. Montipora millepora, sp. n. (Plate XX, figs. 1 and 2 ; Plate XXII, figs. 1, 3, 4, 5.) Specimen B.M. 410 (dredged from Stn. XXIV) bears an extraordinary resemblance to a crust of Millepora, not only in its green-brown colour and the character of its surface, but also through the small, widely separated pore-like calices ; only under a lens do the septa come into view. The section of the corallum is unlike that of any other Montipora yet described. The single specimen, fortunately broken across, forms a crust of an average thickness of 6 mm., near one edge 3 mm., growing on dead and rotten coral, possibly a Cypbastrea. The edges are blunt, usually bent downwards over the edge of the substratum, but at one corner upwards, following the substratum. The surface is smooth, with fine granules, closely placed, and not continuous with any structure seen in the section. The calices are of different sizes and stages of development. The larger are 0-5 mm. across, and have roughened well-developed septa, some fairly regular, others not, without prominent directives ; a few septa of the 2nd cycle present ; they nearly meet, but there is no columella. The calices are raised on conical or roughly cylindrical mounds. Many of those on the surface are oblique, one side slightly raised, in which case the three lower septa are much broader than the upper, the middle one the broadest. The wholly immersed calices are only about 0-25 mm. across, some of them of irregular outline, with indistinct and irregular septa ; or perhaps with only two or three septa, which are, however, indistinct. The section shows a series of hard, glassy, horizontal strata, separated by thin, vertical rods ; at the base these fuse into a solid layer several mm. thick. In only two places, one below this and one above, are there patches of the usual reticulum. It is evident that at a different stage of growth the appearance of the surface would be very different, as the thin rods would stand up above the granulations. This can be seen at a point in the section where a portion of one of the hard layers is exposed. Possibly, as in Cyphastrea and Orbicella (in which a somewhat similar mode of growth occurs), the whole surface would not be at the same stage at the same time. The species resembles Bernard’s species tenuissima and reticulata, which have conical protuberances bearing the larger calices ; certainty is not possible until Bernard’s speci- mens are re-examined, but it is unlikely that Bernard would not have emphasized the peculiarities of this new species if it had been in his hands. Smooth, Branched Montiporae. Montipora prominula, sp. n. (Plate XXIV, fig. 1 ; Plate XXVII, fig. 6.) No. 327 rises from an irregularly-shaped stem 20 mm. x 17 mm., and expands into MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA — CROSSLAND 183 two irregular lobes, apparently formed of fused branchlets (best described by the figure on Plate XXIY). The section shows two layers of coenenchyme, vaguely divided and both spongy. The surface is covered with minute white points, between which it is reticular. There is no trace of foveolation. The principal characters are the large (1-0 mm.) calices with projecting coenenchymal walls, both of which characters are conspicuous to the naked eve ; and the septa, which, represented only by points at the orifice, are seen lower down to be thick, in two cycles and swollen at their inner ends, where the primaries generally meet to form a thick columella. The septa of the 2nd cycle are generally narrower ; but where the columella is exceptionally broad they may join it. The walls are thin, perforate, and end in a row of upright spinules and costal ridges along then1 rims. These walls are not developed on the tips of the branchlets, though appearing immediately on their sides. Xor are they developed round some calices in the grooves between the the main lobes, or they may be reduced to low rings. But over all the sides of the branchlets and ridges they are conspicuous, and usually higher on the proximal sides, thus directing the opening upwards. In growth form this species resembles M. divaricata Briigg. (Bernard, 1897, p. 39, pi. iii. fig. 4), but in that species the calices (0-75 mm.) open flush with the surface and the septa, though stout, are very short. M. compressa Esper. from the Philippines (Bernard, 1897, p. 42, pi. iii, fig. 5) is “a loose bush-like cluster of long and very irregular branches ” but flattens in a somewhat similar way : the coenenchyme is divided sharply into two layers, the outer of which is nearly solid and the calices, 0-75 mm. across, are “ lined by a membranous layer of coenenchyme, which rises slightly above the surface as a thin white ring.” In spite of the difference of form, work on the reef might possibly connect these last two species, though it seems unlikely. Montipora digitata (Dana). (Plate XXVI, fig. 1). 1846. Manopora digitata Dana, p. 508 ; 1849, pi. 48, figs. 1, 1 a-c. 1886. Montipora levis Quelch, pi. 8, figs. 2-2 a. 1897. M. digitata (D) Bernard, p. 47. 1897. M. levis Quelch, Bernard, p. 41, pi. xxxi, fig. 19. 1918. M. levis Vaughan, p. 150, pi. 61, figs. 1, In (no description). The single specimen of this species, B.M. 281 (without Expedition number or locality) is the first of any size to be described, and to this is due the fact that it combines the characters of the two species named, the modes of growth of which seem so different. Both Quelch’s and Vaughan’s specimens are mere scraps, the former with crowded stunted branches, and broken remains of long, slender branches in the background ; Vaughan’s specimen only one branch from a more normal specimen. This specimen from the Great Barrier Reef combines both the cockscombs of M. levis and the long slender branches of M . digitata. The outwardly bent tips of branches which give Dana’s specimen its peculiar appearance are not conspicuous, but two broken branches show them perfectly. The closeness and parallelism of the main vertical branches are also shown in this specimen. Bernard, 1897 (p. 44, under M.fruticoscm says : “ I have no hesitation in separating this from the M. digitata of Dana, in which the apertures are irregular breaks in a surface reticulum ( cf . Zooph., pi. xlviii, fig. 1 c).” Dana gives a rough figure which shows an irregular 184 GREAT BARRIER REEF EXPEDITION but distinct wall. In this specimen the wall is more often quite circular. The septa are irregular in breadth, as in Dana’s figure. Distribution : Banda, Solomons, Fiji. Now G.B.R. Smooth-foveolate, Branched Montiporae. Montipora fruticosa Bernard. (Plate XXIII, fig. 2.) 1897. M. fruticosa Bernard, p. 44 ; pi. iv, fig. 2 ; pi. xxxii, fig. 2. 1918. M. ramosa (part) Yauglian, p. 151, pi. 62, fig. 2. This species is near to M. ramosa , with which it is considered synonymous by Vaughan, but the differences seem constant. The collection contains four specimens, B.M. 193, 277, 287 and 300. The first is G.B.R. No. 8, but no locality is given. It is labelled M. fruticosa, M. divaricata, M. compressa by Matthai, who thereby recognizes its distinctness from M . ramosa, but follows Vaughan with regard to the other two species (reasons against which are given under M. ramosa on p. 186). The differences from M . ramosa are : (1) The “ fruticose,” loosely branching growth-form. There are flattenings and fusions, but no large, long fusions of parallel branches. These specimens, like Bernard’s and Vaughan’s (Plate 62, fig. 2 labelled ramosa), are all small, the largest 11 cm. high. (2) The small size of the calices, which are, however, conspicuous through their depth and openness, “ as if punctured in the almost solid corallum ” (Bernard). In these specimens the calices measure generally 0-5 mm. or even 0-75 mm., not many so little as the 0-35 mm. of Bernard’s specimens. (3) The solidity of the coenenchyme surface, and its spinulation. These speci- mens show the “ almost solid cortical layer” only near the base, but the threads of the reticulum are all thick. The figure on Plate XXIII shows a specimen in which fusion is at a maximum. Distribution : Recorded only by Bernard and Vaughan. From the Great Barrier Reef only. Montipora ramosa Bernard. (Plate XXVI, figs. 3, 4 ; Plate XXVII, fig. 3.) 1897. Bernard, p. 49, pi. v ; pi. xxxii, fig. 3. 1918. Yauglian, p. 150, pi. 62, figs. 1 and 3 (not fig. 2, which is M. fruticosa). 1932. Thiel, p. 114, pi. xvii, figs. 2 and 3. 1938. Eguchi, p. 375. Umbgrove and others on distribution on reefs.* This widely distributed species is known for the variation in its forms of growth. The Great Barrier Reef specimens, though without an Expedition number or note of locality, have evidently been carefully selected, and show a wider variation in form than anything hitherto described, but, paradoxical though it appears, show also that the mode of branching is a specific character. They also show wide variation in structure, com- * Perhaps Crossland was referring to Umbgrove’s (1939, p. 5) work on “ facies-types.” — [A.K.T], MADREP0RARL4, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 185 billing completely smooth with markedly foveolate forms. This is one of those species of great importance on some reefs, but absent, for some quite unknown reason, from others. Umbgrove (1939, p. 12) remarks on its abundance in the Bay of Batavia (in the moats, as at Low Isles), but that it is absent from the Togian Beefs, though carefully searched for. Apparently the specimens were not regarded as of this species by the members of the Expedition, as there is no reference to thin-branched examples in the ecological reports, all mentions of this name referring to M. fossae n. sp. The specimens are B.M. 271, 273, 275, 279a, 279b and 300a. I have added the letters a and b since the two pieces of 279 and 300 do not fit together and those of 300 are of different species. No. B.M. 300a is almost exactly like Bernard's pi. v, fig. 3. Others are more like the very slender form of his fig. 2, which shows the distinct foveolation of the upper ends of the branches characteristic of most of the Great Barrier Beef specimens. No. 273 is of much looser build, long branches with only occasional fusions and com- paratively inconsiderable broadenings of their ends ; it leads on to 279a, in which the fusion of parallel plates is so marked, and 279b, with its palmate branches. This short series illustrates the essential unity of plan in these so different growth forms, viz., the fusion of adjacent branches resulting in plate formation, not merely of isolated branches but of the coralla as wholes, and the broadenings by branchings remaining fused, leading to the palmate ends of 279b. The variation in minute structure is even more interesting. No. 300 is smooth to the naked eye, like Bernard’s figs. 1 and 3, the other specimens showing quite distinct foveolation in places, as in the upper parts of Bernard’s fig. 2, while No. 275 is deeply foveolate all over. Under a lens a slight foveolation can be made out in 300a, the ridges being low and rounded, but here and there thin, with sharp, fringed edges. In one place there is a low mound (Plate XXVII, fig. 4) in perfect continuity with the rest of the branch, and showing no trace of a parasite or other cause of abnormality, in which the foveolae are deep, funnel-shaped and frequently sharp-edged and fringed with spicules. The characters of the calices and coenenchyme differ greatly between the upper, middle and lower parts of the branches. In the upper 1 or 2 cm. the coenenchyme is loose, walls of calices absent or indistinct, septa thin and reaching only to half the radius, as in Bedot’s (1907) fig. 255 on pi. 46 (under the name palmata) and Vaughan’s (1918) pi. 62, fig. la. Lower down, the walls are well made and slightly projecting, septa thicker and broader, and slightly exsert, directives nearly meeting, coenenchyma solid, with low blunt spicules. The difference is greater than that shown in Thiel’s (1932) two figures (pi. xvii, figs. 2 and 3). In the low mound referred to above, the thickening of the walls reaches a maxi- mum, so that there is a space between the calyx-opening and the surrounding ramparts : probably the reason for the sharpness of their edges. This character is frequent in the more deeply foveolate parts of the other specimens ; it is the chief characteristic of M. palmata Bernard (not Dana), which is thus a synonym of ramosa, separated from it by Bernard (p. 66) because it is foveolate, whilst ramosa is glabro-foveolate ; but this series shows that this division cannot, in this species, be maintained. The foveolations in the upper parts of 273 and 279 have an upwardly directed appear- ance to the naked eye, due to the greater development of the ramparts and calyx-walls on their lower sides. They are often broken, so that three calices may lie in a valley, 18G GREAT BARRIER REEF EXPEDITION or a portion of a rampart may be isolated like a pyramidal papilla. This might seem to lead on to the papillae of M. palmata (Dana), but they never resemble those shown in his figure (1849, pi. 44, 2c), in which they are shown as quite isolated and with round tops. Otherwise Dana’s species is remarkably like this. No. 279a is almost completely foveolate, only the main stem near its base being smooth on one side ; the other side is foveolate all over, deeply over the upper half, moderately in the middle area, and not very clearly at the bottom. All the other specimens are smooth at the base and have one side more deeply foveolate than the other. No. 271 is a crust over other corals, probably partly of the same species, very irregular and covered with aborted branchlets from 0-5 to 2 cm. high ; the character of their tops, with fully developed calices, indicates retardation of growth. There are also growing branches at each end of the corallum up to 5 cm. high, and a few between up to 2 cm. Foveolation visible under a lens on the branches, with ramparts romided or sharp, but it is ill developed elsewhere, i.e., over most of the corallum. No. 295 is encrusting a branched coral probably M. fossae ; the crust is thin but knobby, with only rudiments of branches. It is not possible to be certain which of the two species the crust belongs. No. 275 is a collection of branches which I am unable to fit together. Some are much thicker than the preceding, the largest piece being 10 mm. thick and up to 22 mm. wide ; others, on the same piece, are only the usual 5 to 8 mm. thick, and nearly round. The large expansion 22 mm. wide, is not made of fused branches. Synonymy. — It has already been shown that M. palmata Bernard is synonymous with this species. Vaughan (1918, p. 151) gives fruticosa, compressa and divaricata as synonyms, and is followed by Thiel, though, as Thiel says that all his specimens are like the very slender form of Bernard’s fig. 2, they can hardly afford any evidence. For M. fruticosa see under that species ; divaricata is a lobed, smooth form, showing nothing of the characteristic flat fusions of ramosa, and its septa are in two cycles. M. comjpressa is slender and with palmate ends to the branches, and therefore like ramosa in shape, but the projecting thecal walls make promiment rings such as I have not found in any of these specimens, and there is no trace of foveolation. Bernard’s figure on pi. 3, fortunately a collotype, shows these walls clearly under a lens, even on the upper ends of the branches where walls are but slightly developed, if at all, in ramosa. “ As a rule the septa, very irregular and feeble at the margin, become more prominent deep down,” whereas in ramosa they are always level with the wall, and, where better developed, slightly exsert. Esper’s figure of his M. compressa (tab. x, fig. 1) is like B.M. 279b of this series, while his fig. 2 is probably M. fruticosa ; and fig. 3, but for the prominent rings, might also be of M. fruticosa. Another synonym is the foveolate M. indentata Bernard (1897, p. 65, pi. v, fig. 5 ; pi. xxxii, fig. 5). Distribution : Throughout the Malay region to Fiji and Palao. Not recorded from Samoa, Tahaiti or the Bed Sea. In the Indian Ocean, Bamesvaram and Cocos Keeling I. Always recorded as abundant on the shoreward side of coral areas. Montipora fossae, sp. n. (Plate XXVI, fig. 2.) This is the species named M. ramosa in the ecological reports, and which forms so important a constituent of the moat fauna (hence the specific name). It differs from MADEEPORAEIA, HYDROCORALLIXAE, HELIOPORA, AND TUBIPORA— CROSSLAND 187 Bernard’s M. ramosa (three forms are illustrated on liis pi. v) so greatly in its growth form that, in the absence of intermediates, they cannot be united. The species is figured, in its natural surroundings, by Stephenson and others in five photos on pis. ix, x and xvi, all of which agree in showing the same thick, blunt vertical branches with trifling variation, quite evidently the same as Samples 6 and B.M. Nos. 286 and 282. The uniformity of the species shown in these photographs strongly supports mv considering the species as distinct. True, the habitat is a special one, but we have no evidence that this has caused so remarkable a variation from an allied species, especially when we find that other corals of the moat are not particularly affected. The species belongs to the glabro-foveolate group, No. 6 being distinctly foveolate all over. No. 7 has smooth areas, while B.M. 286 and 282 are smooth but for small areas of weak foveolation. These areas in No. 286 are discoverable under a lens, but in 282 are, here and there, visible to the naked eye, as in the other specimens. These two specimens correspond with the photographs above mentioned by Stephenson and S. Manton, No. 286, with pi. ix, fig. 1 on the left, consisting of two upright branches, fused together through the lower halves, and ending in smaller branches, partly fused with the typical blunt conical ends. The fused base is about 34 mm. x 20 mm. ; exact measurements are impossible owing to irregularity and rounded or conical swellings. The free branches are rather flattened, 25 mm. x 20 mm., and 20 mm. x 17 mm. No. B.M. 282 has a thinner stem, 13 mm. thick at the base, where it is dead, broaden- ing and flattening where it is overgrown by living corallum from above, and still more so higher up. It is too irregular for measurement. This main branch appears to have grown at an angle of 45°, and bears 5 vertical branches about 20 mm. thick, ending in blunt cones ; in both specimens these seem to be more slender than those shown in Stephenson’s and S. Manton’ s photographs. Calices small, near the base 0-5 mm. across, distally a little more. The thecal wall is moderately distinct, septa | to § of the radius, but irregular in breadth and often also in thickness ; a pair of directives often more prominent. Six primaries and one or two rudimentary secondaries are present, but one or two primaries may also be narrow. They are continuous plates and may be slightly exsert. Besides being distinctly foveolate the calices of Nos. 6 and 7 are different, 0-75 mm. across, primary septa broader and more regular ; they do not meet centrally but join a columellar mass lying below their inner ends. Directives generally distinguishable, often broad, but not meeting. These specimens might be a variety of M. ramosa, and Nos. 6 and 7 have been so labelled by Matthai, but proof is lacking. Compare Bernard’s pi. v, fig. 1, in which the lower stems are as thick as those of the present specimens, but note that in M. fossae it is the lower parts of the stems that are the thinner. In Bernard’s the upper parts are a cluster of thin branches, which are not conical at their ends. Compare the note on growth form in the Introduction. Eguchi (1938, p. 376) mentions M. palmata (Dana). As he gives M. palmata Bernard as a synonym this may be M. ramosa, but he goes on to say : “ It is easily distinguishable from the typical M. ramosa by having much stout[er] branches, which is usually irregularly united with the neighbouring branches and sometimes forms more or less massive corallum.” This is almost certainly M. fossae. How many other records, e.g., Baker’s* and Umbgrove’s are really this species there is no direct evidence. * This appears to be a reference to Proc. Zool. Soc. London, 1925, pp. 1007-19.— [A. K.T.] 188 GREAT BARRIER REEF EXPEDITION Foveolate Montiporae. Montipora foveolata (Dana). 1897. Bernard, p. 54, pi. vi, fig. 1 ; pi. xxxii, fig. 12. Specimen 303 (outer moat, June Reef) corresponds exactly with Bernard’s descrip- tion and the left-hand figure on pi. vi ; but fig. 12 on pi. xxxii does not show the interior of the calyx, which is deep, the 12 septa (nearly of the same size) running down to the bottom, and thinning below. When there are one or two thicker primaries they are not opposite : i.e., they are not directives. Bernard does not mention the columellar mass, which is generally present, very irregular in shape and size. (The calices of this specimen often contain sand grains, but the real columella can often be made out clearly.) Sometimes it is represented by thickenings of the inner ends of the septa. Distribution : Tongatabu and Fiji ; now G.B.R. Montipora socialis Bernard. 1897. Bernard, p. 56, pi. v, fig. 4. Specimen 318 is, like the preceding, from the outer moat of June Beef. I am doubtful whether this species is really distinct from the preceding, but two small and incomplete specimens afford no evidence for uniting them. According to Bernard the species differ as follows : M. foveolata. M. socialis. Calices large, to 1-5 mm. . . . Calices less than 1 mm. Calices each surrounded by ridges. . Several calices in one valley. Young calices on meeting points Calices as “ swallow nests ” on sides of the ridges. .... of ridges. These two specimens differ distinctly in these respects, but in foveolata there are occasionally two calices surrounded by one ridge, and the difference between young calices on the meeting points of the ridges and on their sides is not always definite. The structure of septa and columella is identical. The fact that Bernard had seven specimens of foveolata and three of socialis is against their identity. Distribution : Great Barrier Beef and the Gloriosa Islands. These are off the north- west end of Madagascar. Foveolate-Papfflate (“ Irregular papillae ”) Montiporae. Montipora venosa (Ehr.) var. angulosa Klz. (Plate XXVI, fig. 5 ; Plate XXVII, fig. 5 ; Plate XXVIII, fig. 7.) 1879. M. verrucosa, Klunzinger, p. 35, pi. vi, fig. 10 ; pi. v, figs. 14, 15 ; pi. x, fig. 7 a. 1897. M. venosa (part) Bernard, p. 69, pi. xxxii, fig. 15. 1906. ? M. venosa Marenzeller, p. 63, pi. 21, figs. 66-68 ; pi. 23, figs. 66a-68a. 1907. not M. venosa Bedot, p. 274, pi. 46, figs. 260-262 ; pi. 47, figs. 263-266. 1918. M. venosa Vaughan, p. 153, pi. 63, fig. 3. 1918. M. venosa Mayer, pi. xix, fig. 46. 1925. M. venosa Hoffmeister, p. 50, pi. vi, figs. 2 a, 2b. The two specimens of this species are remarkably uniform, in contrast to those described above. They correspond well with Klunzinger’s figures mentioned above. MADREPORARIA, H Y DR OCOR ALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 189 Bernard (p. 35) includes Klunzinger's species with tuberculosa. Blit the latter is quite distinct, and, as Marenzeller points out (p. 61), is M. monasteriata F. Bernard’s only figure, a drawing, does not elucidate either species. Bedot's M. venosa is totally different, notably in the looseness of the reticular surface, absence of calicinal walls, and irregularity of the septa. Marenzeller's description and figures are unconvincing, especially in the abundance of thick rounded ridges. Vaughan and Umbgrove accept Bedot's identifica- tion ; Vaughan and Hoffmeister that of Bernard without question ; and none of them refer to Klunzinger. The most interesting feature of the present specimens is the absence of free papillae. Either the calices are level with the surface on smooth areas, or they are surrounded by coenenchymal ramparts which make shallow funnels. The correspondence with Klun- zinger’s figures, pi. v, fig. 15 and pi. vi, fig. 10 (left hand), is exact, except that in the former the rampart round one calyx is divided into 5 papillae, and, in the latter, more or less free papillae can be seen here and there. A point mentioned only by Vaughan is that the secondary septa often bend towards or meet and fuse in a straight “ V ” with an adjacent primary, and there is often a columellar mass. No. B.M. 394 is completely foveolate, except in depressions, where it is smooth. The surface is knobby, and on these knobs and on lower elevations the foveolation is perfect, the ridges being sharp-edged and enclosing a shallow, funnel-shaped depression, at the bottom of winch is the calyx. This specimen was growing on a mangrove-root in Pool 3, Low Island. The root had decayed away except for some of the bark, and the corallum forms a hollow cylinder with edges tinned outwards below, very little free, as the main lamella shown on Plate XXVI follows a branch root ; but at the top the lamella turns over the top of the cylinder and downwards inside it for 1 or 2 cm. A fragment tied to this specimen, but not part of it, is from a basal expansion. Its characters are the same as the corresponding parts in B.M. 394. No. Gr.B.R. 424 is a small crust growing over unidentifiable coral ; its calices are smaller and more numerous, the ramparts form ridges in places, and are, here and there, broken into single papillae, not in a group round a calyx. Otherwise it is like the other fragment. The calices and their variation deserve fuller description. All are 1-0 mm. across, or a little over. Those on the smooth areas are immersed, with up to 12 narrow but very distinct septa, the primaries broader, but a deep open fossa is often left. The directives may be distinguished by being thicker ; or they may be broader also, when one may bear a columellar thickening, or both may join one, or (where the columella is broad) all the primaries may join it. Calices have well-marked walls, which may project as a low ring. Round the calices which are surrounded by the funnel-shaped ramparts the wall is naturally less conspicuous, but it is clearly present. The septa are as above. Even on the upper surface of the cylindrical corallum, where the coenenchyme is loosely reticular, walls, though thin, are distinct, and the septa (also thinner) are straight and well developed. On the foveolate basal expansions some of the fossae are deep and open, and at the edges calices tend to run in rows and the ramparts to form ridges. The coenenchyme is spicular and densely reticular : on the “ back ” of the cylinder nearly solid. It throws up very delicate, branched spicules round the funnel edges, which often broaden into plates ; these, of course, are not comparable to tubercles. G.B.R. 14, a fragment of a crust, is probably of this species. The hummocks of the surface are low, separated by smooth valleys in which are well separated calices about 190 GREAT BARRIER REEF EXPEDITION 0-75 mm. across. These have 6 primary and a few secondary septa, with columellar plugs. Most of the hillocks hear low, pyramidal, separated papillae, but in places they fuse together to form ridges and the usual funnel-like walls round the calices. This specimen is also very close to M. elschneri Vaughan from Fanning Atoll (1918, p. 154, pi. 64, figs. 1, la). Distribution : Red Sea, Northern Celebes, Murray Islands, Samoa ; now G.B.R. Bernard’s record from Fiji is doubtful, though it must occur there. Foveolate-Papillate Montiporae (papillae fusing to ridges). Montipora prolifera Bernard. 1897. Bernard, p. 93, pi. xviii. I identify sample No. 10 with this species, with the proviso that I think it possible that a considerable series of whole coralla might prove its identity with M. foliosa. According to Bernard’s classification the species are widely separated, foliosa being a tuberculate form and prolifera papillate ; but this distinction is not absolute. Of pro- lifera Bernard says that in young fronds the papillae may be quite small, and indistinguish- able from tubercles. In this specimen the surface has overgrown two older growths, the whole being a plate in three layers, so that it is probably young. The tubercles are much more abundant than the papillae, which are so “ proliferous,” and so overgrown by tubercles and plates as to be hardly recognizable as such. The ridges in this specimen are conspicuous only 1 cm. or so from the edge, and on the tops of the larger processes. They are very thin, and less than 1 mm. high. The great majority of the calices are immersed, with only one or two tubercles standing near their walls. Some, on the processes, may be surrounded by tubercles, but the ring is very rarely complete, or regular. Bernard says that there are 12 septa : I find this rare ; 6 plus 3 or 4 is more usual, the secondaries being small, while the primaries may exceed half the radius in breadth. There is often a minute columella-like body. Distribution : Ponape (Caroline Islands) and Amboina (Netherlands E. Indies). Now G.B.R. Montipora undans sp. n. (Plate XXIII, fig. 3 ; Plate XXVIII, fig. 2.) The name indicates a likeness to M. undata Bernard (1897, p. 98, pi. xxi, fig. 2 and pi. xxxiii, fig. 9). This latter figure, a drawing of enlarged calyx and ridges, contradicts the text, as do others of these drawings. Specimens 13 and 187 agree closely; they are small, nearly complete and partly encrusting coralla of irregular outline, saucer-shaped, but with the edges turned downwards ; major diameters 65 and 45 mm. There is no information as to localities of either. The thick continuous ridges rise into crests and meet occasionally as shown in Plate XXIII. The valleys between are much narrower than in M. undata, Bernard’s figure of which, it must be remembered, is reduced to half size. The central area is covered with short plates and round papillae. In the valleys are no small papillae or tubercles, but occasional short plates and thick rounded papillae. The calices are 0-5 mm. across, with clearly circumscribed margins ; septa 6 plus MADREFORARIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 191 2 to 4. rather thick, broad, symmetrical, not meeting. The walls and septa may project a little. Many calices are borne up on smooth processes resembling the papillae to various heights not exceeding 2 mm. ; these are always more or less oblique, even in the central areas, though those only slightly projecting are level, even in the marginal areas, where the tubular ones are procumbent. The coenenchvme is divided sharply into three layers ; the middle a delicate open reticulum bounded above and below by very thin layers ; the upper of very short, close- packed trabeculae ; the lower stony and not distinct from the epitheca. The ends of the trabeculae cover the upper surface of valleys densely, and are like minute upright plates, usually set at right angles to the ridges or radially to papillae. The surfaces of the ridges and papillae are closely granular. There is no reticulum on the surface such as characterizes M. undata , nor is any of the surface spicular as shown in Bernard’s fig. 9, pi. xxxiii. It is tempting to consider these as possibly only young coralla of M. undata. The valleys might broaden out as growth extended, and even the papillae and plates of the central area become continuous serpentine ridges ; but the differences in the calices and their supports, the presence of an epitheca, and the character of the coenenchyme are among the features not likely to change with age. Moutipora sulcata sp. n. (Plate XXVIII, fig. 6 ; Plate XXIX, figs. 2, 5.) Specimen 16 is a curved segment of a circle, measuring 15 cm. radially (the actual centre is unfortunately missing) x 18 cm. across the chord. The right-hand edge looks as if it were broken, but the appearance is due to a sharp downward bend of the straight edge of the plate on this side. The name refers to the long, comparatively regular ridges which cover the upper sur- face from close to the centre to the edge. The characters in which the species differs from other species with somewhat similar ridges are : (1) Absence of epitheca. (2) Height, thinness and continuity of the ridges. (3) Small and very small papillae (tubercles?) between the main ridges. (4) Small, inconspicuous, immersed calices, very few in number ; those on the underside very small. (5) Septa in one cycle, sometimes one or two of the second cycle present. The plate is thin, from 4 to 5 mm. thick centrally, 1-5 mm. at the edge. The ridges are equally well formed, though not so high near the centre of the plate, but here are serrate, showing their origin by fusion of flat papillae. But over most of the corallum they are only slightly sinuous. The ridges are from 2 to 3 mm. high, but only from 0-5 to 0-75 mm. thick below (the upper edge much less), almost quite smooth above under a lens, but their sides are finely spinulose. For details of their junctions and endings see Plate XXVIII. Between them (never upon their sides) lie the calices and small, nearly cylindrical papillae, much smaller than those which, by their fusion, give rise to the ridges ; many are quite minute. These papillae are scattered, but often one stands 192 GREAT BARRIER REEF EXPEDITION immediately proximal to a calyx. The surface of the coenenchyme is rougher than that of the ridges and papillae. All calices are completely immersed, there being no projection of the wall, which is not visible as a distinct structure. They are small, measuring 0-5 mm. and less across, rather far apart, and so inconspicuous that they have to be searched for with a lens. Septa well defined and broad, but second cycle at most represented by two narrow septa. The surface of the underside is smooth (a lens reveals minute blunt spinules) but raised into numerous warts, on one side near the base developing into short root-like processes. Some of the warts bear calices at their points like those of the upper surface, and others may be found on the flat surface. Over most of the corallum, however, the calices are almost obliterated by thickenings of the septa : over the proximal half they can hardly be made out at all. The coenenchyme is excavated by linear, bent pits, the remains of the original reticulation. In section the coenenchyme shows two layers only, a thin solid layer beneath, above which is an irregular network rising into the papillae. The species resembles M. pulcherrima Bernard (1897, p. 91 ; pi. xvii, fig. 2 ; pi. xxxiii, fig. 7) but M. sulcata has (1) no epitheca, (2) continuous ridges, (3) calices all im- mersed, (4) small papillae, or tubercles, between the ridges, besides other differences. Montipora tertia sp. n. (Plate XXV, fig. 2 ; Plate XXVIII, fig. 4.) Specimen No. 11 (no locality) is thus named as it is the third species known in which both the directive septa are highly exsert ; the other two are M. exserta Quelch, the pecu- liarity of which is shown in his pi. viii, fig. 5b (Bernard in pi. xxxi, fig. 13 does not attempt to do so) and M. saxea Bernard (p. 180 in the appendix) from G-ardiner’s Funafuti Collection. M. exserta is glabrous, in M. saxea the septa are only half the width of the radius and irregular in shape, the papillae “ small pimples without any definite association with the calices.” This specimen No. 11 is a half column, evidently grown vertically, formed of an incrustation loosely overgrowing a cavernous mass of the same species, some parts of which are continuous with the living layer ; the hollows and spaces of this mass are due to the mode of growth. The living side of this specimen is made up of irregular vertical branches fused to- gether and bearing numerous abortive branchlets as shown on Plate XXV. The papillae are low and rounded, often directed upwards and elongated in this direction, sometimes forming low rounded ridges. The branches seem to begin as fusions of papillae and ridges. Except on smooth, depressed areas near the base the calices are hardly visible to the naked eye, being only 0-5 mm. across. Seen under a lens directly from above, the 6 primary septa with two broader and thicker directives are conspicuous ; these latter nearly meet at the centre, and are nearly joined by the lateral primaries, but fusion is not visible. Seen slightly obliquely the exsertion of the directives is striking, and some- times some of the others are exsert to a less extent. Some very thin short secondaries are often seen. There is a not very distinct “ petaloid ” wall. Most calices are immersed, the adjacent coenenchyme smooth with minute spicules. Other calices lie at the base of a papilla which forms a hood over it ; or they may open MADREPORARLA, HYDROCORA LLTNAE, HELIOPORA AND TUBIPORA— CROSSLAND 193 on the side or summit of a papilla, or on a ridge. They are well developed on the spongy upper surfaces of the corallimi. A small portion of the under surface is visible ; it is smoothly reticular with small spicules, calices, all immersed, some with thick slightly projecting walls. In another part small, aborted, spicular papillae occur. The vertical, trabecular layer takes up most of the section ; the middle layer is narrow and spongy, in places absent ; the lowest layer may be of short thick vertical trabeculae, or, where the under surface is not living, a thin solid plate. An epitheca is not visible. Foveolate-Papillate Montiporae (papillae regularly nipple-shaped). Montipora verrucosa (Lamk.). Two specimens ; No. 9 identified by Matthai, and No. 12, neither with locality. Both are typical, lamellate in form, with none of the remarkable variation figured by Vaughan (1907, pis. 53 to 59.) Vaughan’s specimens came from Hawaii, a marginal belt island where possibly variation is extreme. Even Bernard’s series was more uniform. Distribution : Cfreat Barrier Reef to Hawaii, including Fanning and Funafuti Atolls ; but not recorded from Samoa or Tahaiti. There is a form very like it in the Red Sea (not verrucosa Klz.), but Bernard thinks this is distinct. Tubereulate Montiporae. Montipora erythraea Marenz. (Plate XXIV, figs. 2, 3, 4 ; PL XXVII, figs. 1, 2.) 1906. Marenzeller, p. 58, pi. 22, figs. 73, 74 ; pi. 23, figs. 73 a, 74 a. 1935. Manton, p. 294, pis. vii, xi. 1939. Umbgrove, p. 55. Marenzeller says, £' Diese Montipora steht der M. foliosa (Pallas) Bernard’s nahe, und ist vielleicht nur eine lokale Form derselben.” The species, after much considera- tion, seem to me quite distinct, and its discovery in the Bay of Batavia and on the Great Barrier Reef corroborates this. The two specimens are widely different. No. 319, from June Reef “ Int. Mad. Zone,” is heavy, humpy and rough, and its surface recalls that of M. prolifera. No. 30 is a thin plate covered with ridges over the whole 10 cm. which it measures radially. No. 319 is heavily wrinkled. In the proximal part it is as much as 18 mm. thick, but at the edge thins down to only 1-5 mm.; it measures 12 cm. radially and 9 cm. across. It is deeply concave below, the hollow 3-5 cm. deep, and correspond- ingly raised into humps and ridges on the upper side as shown in Plate XXIV, fig. 4. The literature later than Marenzeller’s work supports his contention that this irregularity and flatness of the leaves is a distinction from M. foliosa ; but no part of Marenzeller’s specimens was so old and heavy as this, nor so completely covered with ridges as is No. 30. Neither bears upright branches ; but the presence or absence of these is certainly not a specific character in other foliose species. The main distinction from M. foliosa and other species is the way in which many calices are completely surrounded by tubercles. As they increase in length the tubercles carry up the calyx with them, forming tall pillars. These are described by Marenzeller, who duly emphasizes the fact that their ends remain free, fusion taking place only low 194 GREAT BARRIER, REEF EXPEDITION down. All stages in this remarkable growth are shown in this specimen, No. 319. It is quite different from the formation of a calyx on a papilla, or from the compound arrange- ment in M. composita, n. sp. described on p. Ill, or from the fused collars, or imperfect rings of M. foliosa referred to in my note on that species on p. 112. On the distal ridged part of the corallum the pillars are less conspicuous, and generally oblique. As I have remarked before, it is a real misfortune that Marenzeller’s photographs are so coarsely printed, thus making necessary the fresh illustrations on Plates XXIV and XXVII. Calyx walls everywhere distinct ; septa well developed, 6 plus 6 or 4, breadth about half the radius, secondaries narrower, directives not prominent. In this heavy specimen (No. 319) the lowest layer of the coenenchyme is quite solid and measures 1-5 mm. in thickness at the middle of the radius, but 2 mm. nearer the centre. Specimen 303 consists of two fragments, (1) from the edge and (2) from near the edge of a leaf. The pillars are therefore short. Specimen 30 is included here because I can make out no clear difference between it and the distal edge of M. erythraea. but, in his specimen, these conditions hold for the whole 10 cm. which it measures radially. The calices are surrounded by similar tubercles in the same way, but are little, if at all, elevated. The underside is covered to 25 mm. of the edge by lithothamniae and brown weed, more thinly than in No 319, the whole plate being evidently much younger. It is deeply wrinkled, maximum thickness 8 nun., but mostly much thinner, 4 mm. on the average, 2 mm. near the edge. Structure of section as in the preceding. Free part of the underside smooth, without tubercles or papillae ; here and there a rudimentary calyx can doubtfully be made out. It is possible that several large and complete colonies would complete the connection between these specimens and those of Marenzeller, or, on the other hand, show them to be distinct species. In the absence of such material I content myself with a figure which will be useful in the future. Specimen 136 from Traverse II, Deep, is an irregular column, 13 cm. high, 8 cm. wide ; the base encloses a mass of Pontes, and there is little doubt that this gives the shape to an incrustation which is not likely to be anywhere thicker than the section exposed at the base, viz., 4 to 8 mm. As there are no free edges there are no long ridges, but neither are there short ones elsewhere. In No. 136, as in 319, the tubercles are much reduced in the hollows and near the base, where they become small groups of mere spicules. Sometimes in such areas the calyx wall may stand up as a glassy tube independently of the surrounding reduced tubercles. Bare spaces showing a reticular coenenchyme are found here and there. The calices on the sides stand out horizontally, and are not inclined upwards. There is one giant theca, with 12 nearly equal septa, all of which reach a broad flat columellar mass. The hard layer of the coenenchyme is somewhat thinner than in 319 but, as in that specimen, the middle layer is very narrow where it is present at all, and the trabeculae of the upper layer thick and glassy. Distribution : Bed Sea, Bay of Batavia and now G.B.R. only. Montipora foliosa (Pallas) Bernard. This species is recorded by Bernard from several localities between Mauritius and the New Hebrides ; by Vaughan from Cocos Keeling ; by Faustino from the Philippines ; MADREPORARIA, HYDROCORAIUNAE, HELIOPORA AND TITBIPORA— CROSSLAND 195 by Umbgrove from the Bay of Batavia ;* and by Eguchi from Palao where it may reach 3 m. across. It may thus be expected to be common on the Great Barrier Beef, but it has not been recorded from this region except for a doubtful mention by Saville Kent ; and it is not present in this collection. Nevertheless I think S. Manton was probably right in giving the name to many of the foliose forms she met with, though many species are hardly distinguishable until closely examined. I attribute the absence of the species from the collection to the quite unavoidable but regrettable limitation of the packing and transport facilities available in such a place as Low Isles, and to the size of the launches used for expeditions. Most of the older records probably refer to several foliose forms, but we have the following reliable descriptions and figures. Bernard (1897, pp. 156 to 162) describes specimens marked “ a ” to “ w,” and figures “f” on pi. xxx. PI. xxxiv, fig. 13 (a drawing of a single calyx) contradicts the text which says (p. 158): “ they are mostly protuberant, that is, surrounded by rings of tubercles within which they rise.” One tubercle per calyx can be seen in some cases in Vaughan’s (1918) pi. 65, fig. 2a. (Vaughan gives no description.) Bedot’s (1907) pi. 50, fig. 276 is similar : the rings are neither complete nor do they form pillars. Faustino’s figures on pi. 82 show thin tubular calices without distinct tubercles. This shows how distinct is M. foliosa from M. erythraea apart from the marked differences in the arrangement of the ridges, etc. Montipora informis Bernard. 1897. Bernard, p. 133, pi. 27, fig. 3 ; pi. 34, fig. 3.f 1918. Vaughan, pp. 156, 158 ; pi. 64, figs. 3, 4, a-c ; pi. 65, figs. 1, la. Sample 181 from “ Traverse I outside” is a portion of a plate measuring 12 cm. radially and 19 cm. across the chord. It is slightly wrinkled, and bears two small vertical branches, the longer 24 mm. high. It is like the specimens described by Vaughan : rather more like his M. aff. informis than the typical form. In brief the main character of the species is that the tubercles are confined to small groups of 3 to 5, generally proximal to certain of the calices, in very few cases forming a ring round them. Most of the surface is covered with spicules rising from the delicate reticulum of the coenenchyme. This specimen differs from both Bernard’s and Vaughan’s in that the epitheca is confined to a thin line along the edge where it is growing over a dead part of the plate ; the whole under-surface bears very numerous, smaller thecae, 0-5 mm. to 0-75 mm. across, corresponding exactly to Vaughan’s description and figures. There are small scattered protuberances carrying a small proportion of these calices. Those of the upper surface are very nearly 1-0 mm. across, as in Vaughan’s variety. Distribution : Cocos Keeling ; N. Celebes ; Murray Islands ; and now the G.B.R. Montipora composita sp. n. (Plate XXVIII, figs. 1,5; Plate XXIX, figs. 1 , 3, 4.) Two specimens (Nos. 23 and 186) are given this name, in spite of their widely different appearance ; it refers to the arrangement of the tubercles into fused groups, as seen in * But, like M. ramosa, not from the Togian Islands, though carefully searched for there, t This figure shows a ring of widely-spaced tubercles. In this specimen 181, as in Vaughan’s, they are close together ; though not fused. I have ceased to expect accuracy in Bernard’s drawings of calices. 196 GREAT BARRIER REEF EXPEDITION section, which form warts on the upper surface, whence their free ends project. Between these warts single tubercles cover the surface. No. 23 is a young leaf, thin and delicate, 3 mm. thick near the edge (average 4 mm.). Longest radius 13 cm., chord 18 cm. No. 186 is much larger and stouter, 6 mm. thick (peripherally 4 mm., at the edge 3 mm.), measuring radially 21 cm., chord 25 cm. It grows over an older sheet which is 10 mm. thick centrally. In the young piece the warts are low and crowded, the tubercles short, and there is no great difference between those on the warts and those standing alone, so that the surface has a remarkably uniform, smooth appearance. The large specimen, No. 186, is on the other hand rough, the plate itself being wrinkled and hummocky, warts and tubercles higher and less regular ; and the general aspect recalling M . prolifera and M. erythraea. It resembles the latter in having calices raised on “ pillars.” These are completely surrounded by tubercles, the ends of which are free. The tubercles are remarkable for being compound, i.e., composed of other tubercles around and below those immediately surrounding the calices. There is also no formation of plates centrally, or of ridges peripherally, except for a small area near the edge of the big specimen, where they are extremely thin, perforate and jagged : their origin from tubercles is obvious. The warts may be arranged in irregularly circumferential lines in the small specimen, in the larger in no particular direction. No. 186 is infested with numerous small balanids, which have no share in the formation of warts. Though they seem to prefer to settle on warts their effect is to shorten and flatten them ; and uninfested warts are in a great majority. The young leaf, in which the warts are small, is free of all parasites. The tubercles are long and slender, round with rounded tops ; and though covered with fine spicules the general effect is smoothness. A few are larger and flattened — I presume what Berhard calls “ flame-shaped.” Those between the warts are shorter, but like the others so closely placed that little (or, in the smaller specimen, practically nothing) can be seen of the coenenchyme. What can be seen of the surface appears solid in the small specimen : in the larger reticular, with small rounded meshes. Calices numerous ; many not supported, but immersed between the warts, some of which have no relation to tubercles. Many of the calices are oblique (to be seen only by tilting the plate), and may point in any direction over the central area, but in the peripheral part generally towards the edge. They are 1*0 mm. across, with 6 broad primary septa and several narrower secondaries. The primaries may nearly meet, and apparently do meet low down, or there may be a columellar nodule. In some places directives clearly marked ; in others all primaries about the same size. The underside is smooth, solid, with very fine spicules. There are numerous small calices with degenerate septa borne on conical protuberances. In the old specimen the peripheral calices become elongated, in one part up to 5 mm. long, and adpressed to the plate. The young specimen, No. 23, shows no epitheca. In No. 186 it extends to 7 cm. from the edge, with some small patches further out. It is much overgrown, part of it by irregular layers of coral, differing from the rest in its immersed calices. These are crowded, have reticular coenenchyme between, and slightly projecting walls with well developed septa. In spite of all these marked differences there is no doubt that these layers belong to this corallum, as in one place a patch is continuous with the rest of the surface. M. composita may be compared with Bernard’s description of M. aequituberculata MADREPORARIA, HTDROCORALLINAE, HELIOPORA AXD TUBIPOR A— CROSSLAND 197 (p. 130, no figure) : " Single calices or small groups are raised up by the tubercles so as to form small excrescences on the surface.” But this seems hardly equivalent to the wart-formation of my species ; nor can the tubercles, with their rounded tops, be compared to “ erect cylindrical flames.” Montipora angularis* sp. n. (Plate XXV, fig. 3 ; Plate XXVIII, fig. 3.). Specimen 137 is from the anchorage, presumably at Low Isles : No. 5 has no locality but one concludes from its number that it was collected long before the other, making it unlikely that they are parts of the same corallum. Both consist of a thin crust, growing- over dead plates and branches of the same species. This soon gives rise to lobes, plates and thin branches, which fuse again into plates in all kinds of irregular ways. The top of the corallum is composed of free slender pointed branches, 5 to 8 mm. thick. On them are long ridges separating areas of tubercles which give them an angular shape. The ridges are always bent and some are broadened and flattened near their ends. The tubercles, apart from the ridges which show their origin from fusion of longer tubercles quite clearly, are generally cylindrical, of various lengths and thick- nesses, with frequent flattenings and fusions into plates. Many of the shorter tubercles form incomplete rings proximal to the calices ; on the branches these are longer and turn the calyx upwards. Even these tubercles are often broadened, and the direction of the flattening may be either radial or tangential to the calyx. Calices mostly lie in the flat, and generally smooth, valley-bottoms, where tubercles are low or absent, and areas of finely spicular and solid or reticular coenenchyme are left bare. None are to be found within 1 cm. or so of the tips of the branches. They are 0-75 mm. across, with 12 septa ; the primaries reach nearly to the centre and the directives meet, but without fusion ; there is no columella, nor do the septa swell at their inner borders. Secondary septa often turn to join the adjacent primary ones, or they may form a straight “V.'’ The lower part of the corallum (where it forms a thin crust over dead branches) is quite smooth, has no tubercles, is solid looking, and has closely placed spicules. The calices are smaller, placed often at the tops of conical protuberances, their septa irregular and narrower. In passing higher up the corallum one can see these cones gradually changing into groups of tubercles. Sections of branches show a reticular central strand, but the outer layer is very solidly made. In branches low down on the corallum the reticular part becomes small ; in one place in the crust the section is nearly solid throughout, though the fusions of the vertical trabeculae can be made out. The glassy under-layer is thin, and an epitheca is visible in one place. A group of species described by Bernard shows resemblances to M. angularis which differs from efflorescens (p. 150, pi. xxviii, fig. 1) in having pointed, not rounded terminal stems, though the figure shows pointed and looser branchlets lower down the corallum. M. angularis differs in the regularity of its uniform tubercles, which leave no bare areas, and primary septa reach only to half the radius. M. ellisi (p. 151) has shorter rounder * My Latin Dictionary (published 1711) : “ angularis, crooked, having corners.” Both attributes apply to this species. There is no risk of confusion with M. angulata (Lamarck), which is a “ very markedly foveolate ” form (Bernard, p. 63). VI, 3. 14 198 GREAT BARRIER REEF EXPEDITION tubercles which, form the whole surface, and the calices have columellae. M. striata (p. 154) seems closer ; in this the branches rise from a saucer-like plate, but this, in some other species at least, is not always of specific value. One cannot be sure what Bernard means by “ striated.” The drawing of pi. xxxiv, fig. 12 shows longitudinal bars of the reticulum, and pi. xxviii, fig. 3 is too coarsely printed to help in detail. The branches are pointed, but straight, and do not fuse except basally ; they are like those of M. angularis in having no calices on the uppermost 4 or 5 mm. Septa differ in being thin, slightly swollen at their inner margins ; the fossa is clear and deep. Genus Acropora. This is the largest genus of the corals, and the most abundant reef-former. In my experience, as in that of the Great Barrier Reef Expedition, it is the most abundant and varied in clear water along the reef edge, where its species often crowd out nearly all other genera. Vaughan mentions Bernard’s 74 species from this region, but adds that about half of the names are probably synonyms. Mayer’s Murray Island Collection contains only 18 species, of which five, including one new, are not recorded by Bernard. Forty species are described here, of which five are new (a number attributable to the richness of the area and the skill of the collectors) but it is clear that, however much Bernard’s number of 74 species may be reduced by further research, this collection does not contain even all the known species from this area. The fewness of the species common to the Great Barrier Reef and the Red Sea and Indian Ocean is remarkable. In working out these species I have acquired a great respect for Bernard’s work. Before 1893 the great variability of corals with their surroundings was hardly known, or only hinted at. What was known was the fact that among plants every slight ecologic change has a distinct species to fit it, and it was only natural that the same should have been supposed to apply to the corals. Gardiner was the first to emphasize the difference and some of his successors seem to me to have possibly carried the merging of species beyond what is, at present at any rate, capable of proof. Bernard’s work is open to criticism mainly on account of the number of species which are not illustrated, and the absence of enlarged photographs, but it must be remembered that he had to take his own photographs, and seems to have been unable to obtain them from foreign museums. The day when all coral illustrations published will be actual photographs, like Klunzinger’s, is still in the future. The fact that the species are, in my text (not in the index list), arranged in the order Bernard uses, does not imply com- plete acceptance of his classification. Many species fall into related groups, which do not always correspond with the usual subgenera, but attempts by earlier authors to classify all the species have not succeeded. Compare, e.g., A. cancellata, A. clavigera on p. 225. Some species show remarkable adaptations for resistance to surf by flattening out into plates, with the shortening or disappearance of the stem of corymbose forms. As an illustration of this Crossland (1928, p. 723) describes the extreme forms due to the almost complete absence of tide in Tahaiti ; but even here the species are always recogniz- able as derived from the normal forms growing in the lagoon. This is one indication that the extreme variability of many species under changed conditions has its limits, which could be worked out by observations on the reefs. Brook (1893, p. 8) gives a long description of the septa in his introduction, yet, like MADREPORARIA, HYDR OCOR ALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 199 other authors, often omits them from his descriptions of species. In many species they afford little help, but in others the differences are constant and afford clear distinctions, not likely to be affected by external conditions. For instance directives may or may not be broader than the other septa in the axial calices. It is perhaps worth mentioning that the septa, in these and other corals with more or less translucent wails, cannot be properly seen unless the theca is shaded, by the finger or a fragment of opaque material. Brook was the first to discuss the priority of the name Madrepora, concluding that its use was justified in practice, like that of, e.g., Holothuria. The later change to Acropora has no practical justification whatsoever, but, as it has been adopted by later authors, there is now even less advantage in changing back again. Umbgrove (1940, p. 303) gives a list of the species from Batavia provided by Dr. Verwey, in which several of the names used here are regarded as synonyms. As the list is provisional, and, after full consideration, I find that the names I use are, pending Dr. Verwey’s expected monograph of the genus, the correct ones, I treat the synonymy here instead of taking each case separately. The names I use are marked with an asterisk in the list. 1. A. virgata (Dana) = *A . pulchra (Brook), 3 specimens in the Batavia Collection. Brook p. 40. Figure by Dana, pi. 39, figs. 1, la. Brook places the two species under separate divisions, thus : C, p. 38. Prominent radial corallites, ascending , tubular. D, p. 41. Short, erect tubulars unless near apex of a branch. A. virgata shows the ascending tubulars very markedly in Dana’s figure. Brook says “ none are immersed.” In A. pulchra they are equally conspicuous and erect ; and there are sub-immersed and fully immersed corallites among them. Brook had 11 specimens : this collection has 13, with wide variation, but A. virgata is not found among them. 2. *A. aspera (Dana) = *H. hebes, and two other of Dana’s species. 4 specimens. For A. hebes see Brook (p. 128) ; Dana (pi. 35, fig. 5) ; Vaughan (pis. 73 and 74) ; and Hoffmeister (p. 57, pi. 9). The two species are kept distinct by Brook, Dana, Quelch and Gardiner. I agree that they are not easy to separate, but the following characters of A. aspera constantly differentiate it from A. hebes. (1) Irregularity in sizes of radials, giving a rough appearance. (2) Funnel-shaped ” cavities of axials ; septa deep down. (3) Thin, as well as perforated walls ; and pointed shape of radials. (4) Thin, wavy or bent septa of axials. This collection contains 5 specimens of A. hebes, and 6 of A. aspera. Vaughan had several spechnens, and figures Dana’s type. Hoffmeister had “ a very large and excellent suite.” Both consider their variation. 3. A. millepora (Ehr.) = *A . squamosa (Brook) and four other species. 2 specimens only. Brook, p. 116, says that Heteropora millepora Ehr. is not M. millepora Dana : he had seen Ehrenberg’s type in Berlin. There is no figure of this species in existence. On p. 117 Brook repeats this, says that Dana’s type is lost, but takes his species as synonymous 200 GREAT BARRIER REEF EXPEDITION with his own M. squamosa. There is thus complete uncertainty about the name A . millepora and Brook is justified in substituting a new name, squamosa. The species falls into Brook’s class “ A ” ( Radials gutter-shaped, thin walled), which includes M. eonvexa Dana, M. prostrata Dana, M. subulata Dana and M. selago Studer, all of which Verwey regards as synonyms. Brook himself says the first three hardly differ. Dr. Verwey ’s discussion and figures of the types will be interesting, but the name squamosa stands, for the reason given above. The description of these other species, if their names are synonymous, has not been found sufficiently complete for use by subsequent workers. 4. A. hyacinthus (Dana) = *A. patulata and 4 others, including A. hifaria (Brook), A. Jcenti (Brook) and A. capillaris (Klz.) The difficulty here is to know exactly what Dana’s M . hyacinthus is. I am not con- vinced by Hoffmeister’s identification, in spite of his photograph of Dana’s type ; or, perhaps, because of it. A. hifaria almost certainly owes its peculiarity to having been overturned without being killed, and so having become two-faced, like examples I have seen in Tahaiti. But A. kenti shows a difference on its two sides similar to those of normal growths, and the number and size of the branches of the underside, as well as the characters of the radials of the upper, distinguish it at once from both patula and hyacinthus. I remark that Verwey had only two specimens in place of the long series collected by Mayor. As for A. capillaris (Klz.) of the Bed Sea I can see little resemblance. A. hyancinthus, which was so conspicuous in Tahaiti, I have not seen in the Red Sea. 5. A. acervata (Dana) = *A. gemmifera = *A. spectahilis. There are 9 specimens. Vaughan describes variation in A. gemmifera and compares it with the closely related A. scherzeriana, but does not mention A. spectahilis as a possible synonym. Verwey had 9 specimens, Vaughan “ a fair suite ” and the Great Barrier Reef Collection 6. The differences given by Brook seem to be constant. As for M. acervata Dana the figure suggests A. hriiggemanni, but certainty is not possible. It is caespitose, and “ acervate ” at its apices. A. gemmifera and A. spectahilis are, I believe, always massive and corymbose, as described by Brooks but Vaughan figures specimens which appear to be caespitose, though one is certainly morbid, partly overgrown by lithothamnia. Acropora intermedia (Brook). (Plate XXXII, fig. 1.) 1893. Brook, p. 31, pi. 1, fig. c. Specimens 312, 89, 366a, 309, 78, 71, 377a and P3a differ greatly in appearance and mode of growth. Some of them show the lax branching of several species which grow into areas of tangled stems, in which the lax branches from several main stems form a hardly extricable tangle — a tangle in which the lower parts are dead. As is expected, a free-growing corallum exposed to water movements on every side has a more closely branched and bushy form, and the characters of the corallites typical of the upper ends of branches extend further down, or even to the bases of their branches. There are no fusions of branches in any of these 9 specimens. MADRE FOR ARIA, HYDROCORALLIXAE, HELIOPORA AND TLTBIPOR A— CROSSLAND 201 As this is the best represented of this group of species in the collection, and the one with the strongest branches (2 cm. thick), it is probably this and not A. hcbes, which is the main constituent of the “ platforms ” found near the exits from the moats at Low Isles and near the ends of the outer barrier reefs where they dip a little under the surface, as in pi. viii, fig. 3, and pi. xxvi, fig. 4 of Stephenson and others. The former shows bushy growths in the lower left-hand corner on the edge of the main mass, quite probably the same species as the crowded masses behind. These are extremes of growth-form. The intermediates are 377a, apparently a complete corallum ; and 71, a single branch 23 cm. long, alive for 18 cm., and 2 cm. thick at the base and for most of its length. It divides into two main branches 13 cm. long, each of which bears a number of smaller branches. No. 377 a resembles a main branch of No. 78, (Plate XXXII, fig. 1), but is twice as long and has better developed branching. As regards structure all agree with Brook’s description, but his single figure is of a rather deformed, laxly-branched stem, in which the corallites typical of the upper parts of the stems extend but a little downwards, as in No. 366a of this series. The length Brook gives for the radials, 3 mm., holds for these specimens, but in the great majority in No. 366, some branches of 377 and the fragment P3 and in 89 they are confined to the summits of the long branches. The forms of the radials vary, though keeping to the same type. No. 309 : Badials very prominent, erect and thin-walled ; openings oblique ; septa narrow, but 6 + 4 or 6 + 6 in number. Buds on bases are frequent. No. 312 similar, but walls thicker ; openings circular and very little oblique. Septa as above, but usually broader. These tubulars extend practically to the bases of the two branches (16 cm. long), becoming a little shorter, and the septa thicker and meeting. Nos. 71, 78, and 377 are similar. In all these the upper directive is the longer. Nos. 366 and 89, with shorter but thin-walled tubulars, have their openings a little compressed. Septa 6 + 0 in number, but distinct. Upper directive usually the broader but the two may be equal. They may or may not meet. No. P3a is a fragment, a single branch 13 cm. long and 2 cm. thick, prolonged by a side branch for 5 cm. The main branch ends bluntly, and its terminal calyx cannot be distinguished. On the smaller branches, which are not so blunt, the terminal calyx is scarcely exsert. Septa 6 + 6 in number, primaries nearly meeting. There are no tubulars so long as 3 mm. anywhere ; and over the whole of the main stem they are only 1 mm. in length and adpressed, the inner wall visible but fused to the stem, the opening facing upwards. Calices of this type cover the stem thickly and evenly. Immersed ones are rare and inconspicuous. In all the radials the septa are conspicuous, 6 + 6 in number, primaries nearly meeting. The walls are. thick and spongy, except a small portion next to the stem, the surface peculiarly “ woolly.” This specimen is so unlike the others as perhaps to deserve a varietal name. Two specimens are marked by Matthai “ c. f. ph araonis and grandis .” They are separated at once from A. grandis , the radials of which are “ practically without septa,” and while odd pieces of A. pharaonis strongly resemble A. intermedia , if the coralla, and the series of coralla, are considered as wholes the resemblance disappears. I may remark that I shall continue to regard Marenzeller’s account of A. pharaonis with distrust until it has been worked out again on the reef. Distribution : Becorded only from the Maldives, now Great Barrier Beef. 202 GREAT BARRIER REEF EXPEDITION Acropora pacifica (Brook). (Plate XXXI, fig. 2 ; Plate XXXII, fig. 2.) 1893. Brook, p. 39, pi. xxx, fig. b. There are four fragments of this interesting species : No. 270 (Outer Barrier, June Beef, outer ridge tt 2), a large horizontal branch, 25 cm. long, 35 mm. thick at the base, with seven branches curving upwards and three terminal branches nearly horizontal ; No. 345 (Bibbon Beef, outer moat, also marked v 2) essentially similar, 10 cm. long, 33 mm. thick at the base ; No. B.M. 60 (marked P.5, Batt Beef, surf zone, S.W. side of S.E. corner) ; and No. 333 (other indications illegible,* marked <5 2.). The first corresponds best with Brook’s description and the growth form is similar, viz., “ subarborescent, stout and spreading.” All the details of the calices, etc., are the same, but I should add that there is an irregularity in the direction of the openings of the calices, which is visible in Brook’s photograph. The next two specimens show fusions between horizontal branches, amounting to a flat plate in part of No. 345 (PI. XXXI, fig. 2.) The upright branches are more slender, but retain their characteristic proportions. Only in this specimen No. 345 are the axial calices exsert (2 mm.) as in Brook’s specimen : in the other two they are 1 mm. and less in length. No. 333 is a small scrap, consisting of a group of five little branches rising from an encrusting base. Its well exsert axials and more regular radials give it a somewhat different appearance, but the peculiar characters of its details are the same. From these specimens with fused branches it is evident that the long branches are broken from a massive plate. Some whole colonies, which have never been collected, would be very interesting. It is probable that the two specimens drawn diagrammatically by S. Manton on her pi. xv, and provisionally named A. decipiens, really show the growth form of this species, which, so far as I know, is unique. Distribution : According to Brook, China Sea and Samoa. The latter locality may be doubtful, as missionaries used to trade in corals as curios, and it is not mentioned by Hoffmeister. This is the first record from the Great Barrier Beef region. Acropora grandis (Brook). (Plate XXXI, figs. 1, 3, 4). 1893. Brook, p. 42, pi. 1, figs. A and B. Samples 266 from June Beef and 457 (locality not given) correspond well with Brook’s description and fig. a, but the cluster of branches shown in fig. b is not here represented. These two specimens differ considerably, chiefly in the longer radials and more tapering branches of No. 457. The result is that No. 266 is on the whole smooth, whilst No. 457 is rough. The June Beef specimen No. 266 is 41 cm. long, though the base and the ends of the branches are lost. The two main branches are slightly oval in section, at the bottom measuring 30 mm. x 27 mm., and decrease very gradually to the tops. They are slightly bent. The species is rare, and the distinction from others, e.g., A. robusta, needs emphasis. I therefore add to Brook’s description. * Appear to read, “ Hefty outer moat.” — [A.K.T.] MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 203 The main characters are : (1 ) Size and thickness of the stems, with which is correlated ; (2) a solidity which is complete at the base, and remarkable even in the breaks a few centimetres from the ends of the branches ; (3) a surface reticulation so coarse as to be visible to the naked eye ; (4) the curious grouping of the radials ; (5) radial calices are short, cylindrical or slightly compressed, and very thin walled, of irregidar lengths and opening in various directions ; (6) the blunt ends of the branches, on which the apicals are small and hardly distinguishable from the radials (in No. 266) ; (7) almost complete absence of septa from .the radials. The larger specimen, No. 266, probably lay near the ground, since the longer branches all rise in one direction, though some stumps rise near the base on the supposed lower side, and the calices on this side are shorter and less crowded than on the other. On both sides they become much shorter basally than distally, as in all species, but here they project only about 0-5 mm. for the lower two-thirds of the main stems. In the four stumpy branches of the base, and in one 24 cm. above and pointing upwards, the axial calyx cannot be distinguished ; probably these branches ceased growth some time ago. In the remaining few ends it is just distinguishable. For the curious irregularities of the radials I refer to the figures. A few near the ends of the branches attain a length of from 2-5 to 3 mm. and give some appearance of regularity ; but even here there are shorter and immersed calices, irregular in size and direction. These downward or sideways projecting tubulars are not mentioned by Brook, but can be made out in his figure. Similarly another peculiarity, the placing of the calices in groups or lines of three to six, in which their walls actually fuse, can be seen under a lens in Brook’s figure. This grouping leaves small areas of coenenchyme free, exposing its characteristic coarse reticulation. The costal ridges are thin, but broad and continuous, not spinulate, extending on to the lip of the theca. They broaden as they pass down- wards, so that the corallite appears slightly conical, but do not spread out on to the coenenchyme. Such spinules on the coenenchyme as there are, are broad and blunt : there are none on the costae. The smaller specimen, No. 457, is only 19 cm. long. The apicals of 7 branclilets are present, those of the three main branches being lost ; though no larger than the adjacent radials they are distinct. These young branches are more slender and more pointed than those of the large specimen, though still blunt. In addition there are more pointed branclilets and longer tubulars, aborted branclilets, along one side of each main branch. These, and the irregular lengths of the same elsewhere give the corallum a very rough appearance, compared with which No. 266 is smooth. Otherwise the structures of both are the same. Distribution : Great Barrier Beef only. Saville Kent collected five specimens ; not seen again till now. Acropora pulchra (Brook). 1893. Brook, p. 44, pi. 28, figs, a, b, c. 1918. Vaughan, p. 162, pi. 66, figs. 1, 2, 3, 3a. The collection contains 13 specimens, of which no two are even approximately alike. Evidently they have been carefully selected to illustrate the variation. Eight specimens have been labelled by Matthai, one each as variations alveolata and stricta, and four others with a ? mark. One is not named. The usual colour is given by Stephenson and others 204 GREAT BARRIER REEF EXPEDITION as brown, with pale blue tips ; but No. 217 : “ 0. outside Low Isles, yellow tipped.” This is an extreme variety, but colour may or may not be a specific distinction, as noted in the introduction, each case should be considered separately. Only one specimen (No. 370, Lizard Island, A. Reef) corresponds exactly with Brook’s description and his and Vaughan’s figures (though Brook’s is a mere fragment and he gives no enlarged photographs.) It is a curved stem, 23 cm. long showing numerous round, open and nearly immersed calices, especially in the lower half. Here also, in some cases, septa are fairly developed, whereas in the upper parts one directive alone, the upper, is usually conspicuous, the other septa rudimentary, as in the type. The calices are ridged outside, the ridges passing into a reticular coenenchyme, broken up on the surface into “ dot and dash ” granules. At the base of the stem this granulation becomes more regular, but never fine and even, as in the form described later. Acropora pulchra var. alveolata Brook. No. 91, thus labelled by Matthai, is like the type in having branches that ascend at about 30°, but differs in its “ nariform or half tubular ” radial calices ; these are much less like those of the type in Vaughan’s figure than of this specimen. Brook also says that the inner directive is often broader than the outer, which is very rarely the case here. In short, the characters upon which Brook founds his variety are variable. Acropora pulchra, var. stricta (Brook). (Plate XXXIV, fig. 2.) No. 92 is thus named by Matthai and with it I consider No. 280, from the reef crest, at the north end of June Reef. They differ from all the other specimens in being more frequently branched, the branches thinner and set at a wide angle with the main stems. They are “ more slender and tapering ” but the larger calices are not more distant : in fact in No. 92 they are rather crowded. “ The septa are better developed in the radial corallites ” in some cases ; in others no better developed than in the type. The larger radials in No. 280 often run in rows. They are nariform, as in var. alveolata, but a few on the backs of the branches are tubular. The differences between the variety and the type are probably ecological (see Mayor’s experiments given by Hoffmeister, and here quoted under A. liebes). Acropora abrotanoides (Lamk.) 1893". Brook, p. 56 (no figure). 1918. Vaughan, p. 166, pi. 68, figs. 1, la, 2. Specimen No. 286 corresponds with Brook’s description and Vaughan’s figures, the only ones existing. I add that the axial septa, though broad, are thin, and that the secondaries tend to bend and fuse to primaries ; of these septa, the directives and some- times two others meet. Radial directives may meet deep down, with a thickening at the junction. Immersed calices with a low, ring-shaped wall ; septa represented by directives (which may meet) and rudiments of others, or all may be rudimentary. Axials with small buds near apex ; long tubulars with small buds, or forming short branchlets. Specimen B.M. 293 (no G.B.R. locality) consists of two thick stems, one now lost rising from a broad incrustation, 4-5 x 5 cm. on one side, and about 4x3 cm. on the MADREPORARIA, H YDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 205 other. The remaining branch is 25 mm. thick — the lost one 22 mm. — and 60 mm. high. Half-way up it gives off three branches and some small stumps. The specimen is evidently a surf-swept dwarf. The interesting part is the incrustation: this bears 3 stumps, 1 cm. high and thick, but elsewhere is closely covered with thick-walled, short, tubular calices, generally 2 mm. thick, but varying to small and nearly immersed. Here and there can be made out indistinctly little rosettes, in which an axial is surrounded by, or bears, a few irregularly placed radials. All these calices, even the asymmetrical ones, but not the very small immersed ones, in which the septa are very narrow, have well developed septa meeting a columellar mass. Septa in the calices of the branches may meet, but there is no columellar mass. The species is near to A. danai and A. decipiens ; but Vaughan finds it distinct, as did Brook. Distribution : Singapore ; Great Barrier Reef ; Tahaiti. Acropora affinis (Brook). (Plate XXXIV, fig. 1.) 1893. Brook, p. 60, pi. xxviii, fig. f. This species, though very distinct and conspicuous, is recorded only by Brook. This is true also of his allied species A. gravida , which is doubtfully the same as A. brachyclados (Ortman). The collection contains two remarkable specimens, of special interest, since all Brook’s nine specimens are either very young or only terminal branches. The Great Barrier Reef Expedition’s specimens are massive branches ; No. 356 (without locality) 50 cm. long, 6 to 8 cm. thick ; No. 384 (from June Reef) 43 cm. long 8 cm. thick near the base, about 6 cm. over most of the rest. These branches are irregular both in direction and thickness ; in No 384 the lower 6 cm. is decayed, and only 3-5 cm. thick at the bottom. It seems likely that these are branches from large growths which must have been very conspicuous. A peculiar feature of both, most marked in No. 384, is that forks of the main branches grow parallel and close together without fusing before diverging, so that the fork is divided by a deep and narrow slit. Probably because these are fully grown specimens there are some divergences from Brook’s description. In No. 356, but not in No. 384, the branclilets of the main stem, but not those near the apices, are often in groups of 3 to 5 ; or the branclilet thickens, and bears branclilets of nearly equal size at its base. Secondly, both in axials and radials of both specimens, the septa on the twigs on the main stems and those near the apices are quite well developed, though narrow, as in A. gravida. In all other respects, however, both correspond exactly with A. affinis, as described by Brook. Distribution : Great Barrier Reef and Macclesfield Bank only. Acropora aspera (Dana). (Plate XXXIII, figs. 2 and 3.) 1848. Dana, p. 468, pi. 38, fig. 1. 1893. Brook, p. 62. 1886. Quelcli, p. 156. The six specimens, Nos. G.B.R. 285, 66, 304 and B.M. 299, and two unlabelled, are 206 GREAT BARRIER REEF EXPEDITION low bushy growths, probably from shallow water (localities are not given). No. 304 is a remarkable contrast, consisting of long branches, more gently tapering, evidently grown under more free conditions. This is from the outer barrier of June Reef, presumably from the “ madrepore field ” where the end of the reef dips into deeper water. It affords an exact parallel to the experiment Mayor made in Samoa with A. hebes. Dana’s is the only figure, but he was interested only in the expanded polyps : the branchlet showing the skeleton was merely sketched in. It is therefore advisable to expand the description, particularly of the septa. The blunt-ended branches are given as characteristic, but, as noted, this is not invariable. Besides in No. 304 there are some tapering branches in Nos. 0, 1* and 285, in which most are typically blunt-ended. The thin-walled, generally fragile (Dana’s term), short, rounded, labellate radials are characteristic. They are never tubular, or even partly so, the upper wall being at most a mere ridge, though in No. 304 they are longer, and some have the upper wall distinct. Many are small and sub-immersed, on the lower parts completely so. A peculiarity of these smaller radials, not hitherto noted is that they are often placed in close relation to the bases of the larger calices, and in many cases a little way up their walls, but usually one only is so related to, or budded from, a larger radial. The axials are described by Brook (in Quelch’s specimen) as “ often with a large deep funnel-shaped cup.” I find that though a few are almost without septa they generally have six primaries and a variable number of secondaries. No. 66 e.g., has axial septa as follows ; they are all very thin, and often bent. Axial calices. 5 2 1 2 2 2 Septa. 6 primaries, meeting. 6 primaries, meeting ; also some rudimentaries. 6 quite regular primaries, meeting ; and 6 smaller. 6 primaries and 10 others, all very narrow. Too irregular to count. 7 irregular. In No. 304 the arrangement of septa is : — Calices. Septa. 3 Quite irregular. 1 8 long and 6 short, the former with irregular fusions. 1 6 plus 6, all short. 1 6 plus 0, all short. 1 4 short and 2 directives. In No. 285 the septa are irregular and very thin, often so deep down as to be hard to see. In a few calices they number 6 plus 6, with primaries meeting ; but generally none meet. Dana says (apparently of the radial calices), “ star very short rayed, the directive septa prominent.” In No. 285 most radials appear empty, or with narrow directives, but low down the larger calices can be seen irregular, not countable septa. Sometimes * An unlabelled fragment. [A.K.T.] MADREPORARIA, HYDROCORALLIXAE. HELIOPORA AND TUBIPORA— CROSSLAXD 207 there are very narrow ridges at the tops of the cups. The immersed calices of the lower parts of the specimen generally show very narrow septa ; hut they are irregular in size and number. In other respects, e.g., character of the coenenchyme, the specimens agree with Dana’s and Brook’s descriptions. Distribution : (as given by Brook) : Philippines ; New Hanover ; Great Barrier Beef ; Fiji ; Botuma. Acropora digitifera (Dana). (Plate XXXV, fig. 2.) 1846. Dana Madrepora digitifera, p. 454. 1893. Brook M. digitifera, p. 75. 1902a. Yerrill Acropora digitifera, p. 228 ; pi. 36, fig. 12 ; pi. 36b, fig. 3. 1918. Vaughan Acropora ( Tylopora ) digitifera (Dana), p. 175, pi. 76, figs. 1, 1(7, 2. 1918. Mayer, pi. 13, fig. 7. This species has been thoroughly described by the above authors. Neither Dana nor Brook give figures, but Verrill photographs portions of the type, correcting Dana as to the thickness of the axials. Vaughan and Mayor figure Murray Island specimens. Vaughan’s fig. 1 shows a more straggling bush, with rather more tapering branches than the specimen here, which is more like Mayer’s figure. The present specimen, No. 308 (” Lizard Island, 3.6.29. A. Beef. Mad. L."') is a compact bush, with stout upright branches, 15 mm. thick and 7 to 8 cm. long, correspond- ing exactly to descriptions except that the immersed calices among the ordinary radials are few, and generally not completely immersed. It is labelled by Matthai “ Acropora gemmifera (Brook). This is really A. digitifera (Dana) ” In spite of the close resemblance of the upper parts of the branches above the characteristic lateral branchlets to those of A. gemmifera, the different form of growth alone is a real specific distinction from that species. This view is supported by Stephenson and others’ photograph on pi. 24, showing numerous specimens of A. spectabilis (or gemmifera ?) growing in sheltered water behind the edge of Yonge Beef, none of which show a bushy growth. By contrast S. Manton’s pis. xiv and xv show both corymbose and bushy specimens of “A. gemmifera ” scattered over exactly the same habitat, proving that the different forms are not due to any influence of the habitat. The bushy forms are in fact, a distinct species, vis., A. canalis. Distribution : Great Barrier Beef (Murray to Capricorn Islands), Madagascar ; not recorded east of the Great Barrier Beef. Acropora haimei M. E. and H. (Plate XXXIII, fig. 1 ; Plate XXXV, fig. 1.) 1860. Milne Edwards and Haime, p. 151. 1879. Klunzinger, p. 21, pi. i, fig. 9 ; pi. v, fig. 4 ; pi. ix, fig. 16. 1893. Brook, p. 77. 1906. Marenzeller, p. 51, pi. xvi, figs. 45-48. 1918. Vaughan, p. 164, pi. 66, figs. 4, 5. Klunzinger regards his identification of the Bed Sea specimens with Milne Edward and Haime’s three-line description as safe. Brook mentions the type, but says nothing of the “ orifice ovalaire ou meme presque lineaire ” : he only says that in some specimens it is circular on the edge of the colony. None of the three, nor Marenzeller, say anything 208 GREAT BARRIER REEF EXPEDITION about the septa. Kluu zinger’s figure shows a surf-swept dwarf, though he says, “ Die Kolonie hat verscheidenen Habitus ; bald ist sie (das haufigste) rasen oder rosetten-formig und meist viel sprossig. . . . ” That the turf-like forms are the most common is what would be expected at Koseir : they are not common at other places where there is shelter, more bushy growths are illustrated by Marenzeller, who, as usual, gives no reason for his inclusion of these widely differing forms in one species. (His fig. 46 shows a dwarf form again, but it is recorded from Sherm Sheikh, Sinai, which is completely landlocked. Probably it is really from outside the Sherm.) In Vaughan’s specimens, both of the species and of its variety, the radials are “ very slightly or not at all compressed.” He mentions the presence of two complete cycles of septa in the radials, which I find a some- what striking feature of the Great Barrier Reef specimens. These two specimens, named by Matthai, Nos. 296 and 234, are, to the naked eye, as different as two species could well be, both in growth form and in the regularity versus irregularity of calices and branches. No. 234, labelled “ T2 shallow Bush,” is a single stem 18 cm. long and 10 mm. thick throughout its length, with branches at angles of 15° to 30°. It is most like Vaughan’s pi. 66, fig. 4, while No. 296 rather resembles pi. xvi, fig. 48 of Marenzeller. As the differences are so very striking and resemblances to published figures not complete I give a figure on Plate XXXV, and remark that I cannot feel satisfied with these determinations without having the evidence which Marenzeller withholds, and that to be gained by further work on the reefs. Distribution : Apparently common in the Red Sea ; but, in spite of its wide distri- bution, rare elsewhere. It is not recorded from Samoa, Malaysia or the Philippines, nor from the Great Barrier Reef until 1918. A. tubigera (Horn). 1902a. Verrill, p. 239, pis. 36, 36a and 36f. 1893. Brook, p. 79. Specimen 65 is complete, spreading horizontally from a base on one side, and is 17 cm. long, 18 cm. wide, and 9 cm. high. This specimen is not translucent, as were Verrill’s Singapore specimens, but the coenenchyme is hard and strong, though broken branches show circles of spaces radiating from the distinct, imperforate wall of the axial calyx. “ The main branches . branch dichotomously.” As there is no figure of the whole corallum I do not feel sure that “ dichotomous ” is her# used in its strict sense ; if it were the corallum would be I think, unique in the genus. That “ the distal calices are about as large as the axial, or even larger,” is another apparently abnormal state of things, but the figures show nothing unexpected. Otherwise the description is thorough and good. I am doubtful whether the distinction Verrill makes between his specimens and Brook’s is valid. This one seems to be intermediate in the solidity of the coenenchyme, and there is little difference between a short, adpressed, tubular radial and a round, nariform one. The inconspicuousness (not absence) of the second cycle of septa in the radials is, in this species, a minor detail, since, though most radials show them distinctly, in some others all the septa are much reduced. Distribution : (including Brook’s specimens) : Singapore to Louisiade Archipelago, (just east of Papua). Now Great Barrier Reef. MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 209 Acropora quelchi (Brook). (Plate XXXVI, figs. 1-3.) 1893. Brook, p. 90, pi. 32, figs. d. e. 1907. Bedot, p. 256, pi. 41, figs. 225 to 234. 1932. Thiel, p. 119, pi. 14, fig. 3. Quelcli (under tlie name M. effusci), Hoffmeister and Faustino also record the species, but without description or figures. The species is not very easily distinguished from A. loripes except by growth-form though Brook places them in different sub-genera ( Polystachys and Conocyathus) ; but these are separated only by details of the axial calices, which are often variable, as in, e.g.. A. rosaria. A. quelchi is “ half vasiform from a lateral stalk.” Only one of this series of 9 speci- mens, No. 272, is complete, and this is more bushy than vasiform, (Plate XXXVI, fig. 3), but it evidently grew on the side of a vertical support, its branches bending upwards at an apparent angle of 80° with the horizontal. At the other extreme are specimens which are clearly portions of flattened horizontal growths in which long and compara- tively simple branches are connected by flattened fusions, the ends of which turn up at an anole of 30° or so. No. 305 is intermediate between these extremes, but is much nearer to the bushy form, No. 272. I divide the series into four forms ; they may, in fact, be distinct species, but only evidence from the reef can decide. Almost all are like var. paradoxa Brook (which Brook thought might be a separate species) in having “ swollen hemicotyloid ” instead of wart- like theca among the adpressed tubulars of the lower parts. But I find no axial rods formed by fusion of primary septa in immersed calices in any specimen ; in fact none are perfectly immersed, all having more or less projecting rims. Form A. — A more or less prostrate bush, with numerous long, upwardly bending branches, bearing few lateral branches, and those very short. Long, spreading radials, in distinct rows, over most of the upright parts. In specimen 272 the branches are from 5 to 7 cm. long, tapering, and with fairly prominent axials projecting from 2 to 2-5 mm., 2-5 mm. thick externally, with cavity 1 mm. deep. No. 305 is distinctly of the corymbose form, very like Thiel’s figure ; branches shorter from 4 to 5 cm., and reaching to the same level above ; underside with numerous short branches, the ends of which are more or less horizontal (more so than in 272). Axials exsert only 1-5 mm., i.e., “scarcely prominent” as Brook says for his var. paradoxa , 3 mm. thick, cavities 1 mm. deep. The axials differ from those in A. quelchi as described by Brook, both specimens having only 6 primary septa, as in Thiel’s specimen, rarely with from 1 to 3 rudime ntary septa of the second cycle ; but the primaries do not meet, and are “ subequal and not very prominent.” The septa of the radial calices usually consists of directives, and narrow, almost rudimentary laterals formed of rows of spines ; sometimes some o f the second cycle also present Coenenchyme finely echinulate, everywhere in rows. Form B. — Specimen 279 ; growth form resembling No. 305, but branches longer in proportion, and the development of branchlets on the lower side much less marked. The whole specimen is very light and fragile ; horizontal branches 12 mm. thick, those bending 210 GREAT BARRIER REEF EXPEDITION towards the vertical only from 7 to 8 mm. at base and 5 mm. at a point 2 cm. below their finely tapering ends. Axials 2-5 mm. exsert, 2-5 mm. in external diameter ; owing to the long tapering branches they appear much more prominent than in any other specimen. Their septa number 6 plus 6, are thin and often bent, and may or may not meet, and directives are distinguishable. Eadials thin-walled, much compressed and lips not rounded ; their openings long and narrow, but tubulars with short openings are also common. Septa well developed, the second cycle well represented. Eadials stand well out from the stems and are in distinct rows. There are a few incompletely adpressed tubulars with round openings at the bases of the stems, but no hemicotyloids, and the rims of the subimmersed calices project well. No. Z3 comes nearest to this specimen in delicacy of build. Like all the others, except No. 272, it is unfortunately only a fragment. The underside is distinguishable by a few short horizontal branchlets. The main stems are short and give rise to from 1 to 4 terminal branches, which are straight or only very slightly curved, so that the form is not markedly corymbose. Branches taper, from 8 to 10 mm. thick at base, from 3 to 5 mm. above. Axials 2-5 mm. thick, only 1 mm. exsert, and so not at all prominent ; cavities 1 mm. deep. Their septa generally number 6 plus 4, occasionally 6 plus 1 to 3, none meeting ; directives not distinguishable. Eadials long, spreading and in distinct rows, and mostly distinctly tubular with short oblique openings. Their septa well developed, though laterals may be rows of spines, usually 6 plus 0, but there may be from one to three secondaries ; the outer directives are the larger. In the lower calices the arrangement is the same, but directives not prominent in the subimmersed calices, some of which appear almost empty. I have found only one “ swollen hemicotyloid ” in this piece, and, even on bases of branches, the shorter tubulars are not completely adpressed and the free tubulars are to be found almost to the bottom of the colony. These two specimens, though differing so much, seem to agree on the whole. I keep them as a form of A. quelchi out of deference to Matthai’s labelling, being myself inclined to regard them as distinct species. Only work on the reef can decide. No. P.4 may come here, but there is not enough of it to show the mode of growth. It has completely immersed thecae basally, and for some way up the branches which are practically straight. Form C. — includes Nos. 34, 35, 80, and 284a?. They are all distinguished by their thick horizontal branches, crowded with adpressed or partly free tubular thecae, which, in No. 284a? are mostly “ swollen hemicotyloids.” This specimen is from June Eeef anchorage, and is much larger than the others, which have no locality record and so may be presumed to be from Low Isles : They have none of the broader, basal stems. In none of the four are any of the branches nearly vertical, the main stems being bare almost to their ends. The undersides of 35 and 284a? bear blunt, abortive, horizontal and flattened branchlets with fusions, thus approaching the completely corymbose form much more closely than do the preceding forms. The same structures are less marked in Nos. 34 and 80, but they may have been equally well formed in other parts of the coralla. On the uppersides near their ends the horizontal branches bear a few slightly raised branches, only a little thinner than themselves, and somewhat blunt. MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIP OR A— CROSSLAND 211 The axials of the first three are 2-5 mm. in outside diameter, and exsert from 1 to 2 mm., but No. 284 has axials 3 mm. in diameter, exsert only 1 mm., the internal diameter of all calices measures 1 mm. as usual. Their septa number 6 plus 6, secondaries being very thin. The radials are thick-walled (especially the lower lip) ascending, tubular, and not clearly in rows. Mouths on upper radials compressed oblique, well enclosed by the swollen lip ; round or oval in the lower radials ; some half-free tubulars occur among the compressed mouths low down the stems in No. 35 (the smallest fragment, only 55 mm. long, so that the real basal stems are missing) but hemicotyloids occur in all. The septa number 6 plus 0, sometimes with two of the second cycle. Laterals all narrow, formed of rows of spines. Coenenchyme well spiculated, usually in lines, and thecae striate through- out their lengths. Owing to the great variation in growth-forms according to habitat, there has arisen a tendency to regard growth form, in all cases, as of little specific meaning. This tendency may have been exaggerated, and I am inclined to regard it here, along with other details, as quite possibly sufficient grounds for specific distinction. Form D. — Three specimens, Nos. 70, 81 and 101, are fragments distinguished by their thick blunt branches and short and greatly thickened radials, though the axials are of the usual proportions. This general thickening involves the septa, which are thicker and longer than usual, the primaries of the axials and the directives of the radials almost meeting. In the absence of ecologic or other information I do not feel any certainty that these fragments are really of this species ; two of them are labelled A. quelchi by Matthai. Distribution : Amboina, Banda, Solomon, Philippine and Samoan Islands. Now Great Barrier Reef. Acropora corymbosa (Lmk.). Specimen No. 375 (June Reef, Patch No. 1, square 12) corresponds well with Klunzin- ger’s description, but the axials and the tubulars of the underside are longer, 4 to 8 mm. and 8 to 10 mm. respectively. It is most like his fig. 2b, Taf. ii, but this specimen is of regular shape, and the ends of the branches on the upper surface all of the same length. It does not correspond well with any of Marenzeller’s figures, or those of Hoffmeister, Vaughan or Thiel. The large number of twigs and tubular axials on the sides of the branches near their summits seem to indicate that it is a young specimen which would have grown into a corymbose shape. The septa are described only by Brook ; this specimen corresponds. The shapes of the radials are as the outlines of Klunzinger’s Taf. ix, fig. 19, except that the curious form “ c ” has the narrow end distal, not proximal as in Klunzinger’s sketch. Otherwise they are so similar and so striking that the difference is evidently due to a slip on Klunzinger’s part. Marenzeller gives few details of his specimens, and, as in some others of his determin- ations, I consider that his work needs confirmation — work best done on the reefs. Distribution : One of the few species found throughout the whole Indo-Pacific, including its extremes, the Red Sea and the Tuamotu Atolls. 212 GREAT BARRIER REEF EXPEDITION Acropora armata (Brook). (Plate XXXVII, figs. 1 and 3.) 1893. Brook, p. 100, pi. 10, figs, a, b. Two specimens are both labelled A. delicatula by Matthai, and I have difficulty in distinguishing these species by following Brook’s descriptions, the main difference being the form of growth. A . delicatula is bushy, without a stem, and with long curving branches of irregular length. Should further research bring these species together (and, from what is said on development under A. Tiyacinthus, this is not very likely) the name armata would have the priority as it was the first described species. Only specimen No. 238 is complete with the stem. To it the plate-like portion is eccentric and oblique, corresponding with Brook’s figure and description. I may add that the tubular axial calices thereon are up to 4 mm. thick, those of the underside of the oblique branches at most 2-5 mm., while those of the upperside measure only 2 mm. Their openings measure 2 mm., 1-5 mm., and less than 1 mm. respectively. The stem of No. 76 is unfortunately missing, but it seems to have been oblique. I regard specimen No. 76 as belonging to the same species as No. 238, though it differs markedly in several ways : (1) In general appearance of the upperside the axials long, slender and exsert in No. 238 are not prominent in No. 76, but it is to be noted that these axials are not so much exsert as they appear, as all bear minute buds 1 mm. or less from their apices. In a few cases those of No. 76 are exactly similar. (2) A much more important difference is in the upright branches, which, in No. 238 are often 1 cm. thick at the base, thickening still further higher up (Plate XXXVII, fig. 3), where they divide into from two to five very slender tapering branchlets. The appearance of these thick bases suggests fusion of a group of long slender branches such as those of No. 76. (3) The thicker, downwardly directed, tubular axials and abortive branchlets of the underside are much longer in No. 76, being from 7 to 10 mm in length. The septa of the axials are not mentioned by Brook for either species ; in No. 238 there are two cycles nearly complete, but in No. 76 only the primaries ; in both, primaries are about half the radius wide, and one directive is broader. In other details, such as the very distinct straight costulae, correspondence with Brook’s description is complete. The inturned lips of the upper radials are characteristic. Seen from the side they give the radials a claw-like outline. Brook had eight specimens, apparently all in good agreement. Distribution : Singapore, Diego Garcia (Indian Ocean), ? Tahaiti. Now Great Barrier Beef. Acropora armata Brook var. (Plate XXXVII, fig. 2.) Certain small bushy specimens are referred to by Stephenson and others on p. 83, under the name A. delicatula as follows : “ Apart from these more or less solid forms of Acropora on the outer ridge of Yonge Beef a totally different species occurring particu- larly on the sides of clefts, makes small rounded bushes of branches so slender and brittle MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 213 that an entire specimen can with difficulty be obtained — yet this form can withstand the breakers." Three of the four specimens are from June Reef, Nos. 335 and 359 from the outer moat, and 326 from the outer ridge ; but there is no doubt that these are the same as those referred to. They are all labelled A. clelicatula by Matthai, and it is certainly likely that they are stunted specimens, modified by the environment. At the same time there are differences not likely to be due to conditions, and it is quite possible that more than one species would be modified in the same way, and so converge by simplification. Besides the fact that the branches rise more or less vertically from an encrusting base, with no trace of a stem or main horizontal branches (though there are long tubular calices and abortive branchlets round the edge of No. 326), the axial calices are thick- walled and eostulate, costulae forming only fine lines in the ordinary way. The openings of all calices measure 1 mm. in diameter ; but externally calices are 3 mm., 2-52 mm. and 2-3 mm. thick in Nos. 326, 359 and 335 respectively, the thicker axials corresponding to rather thicker branches. The septa differ in No. 326 numbering in the axials 6 plus 1 to 3 ; in Nos. 359 and 335 there are often 6 secondaries. In all, the directives meet low down and are nearly joined by the lateral primaries. The radial calices are fairly thick-walled, lamellate or nariform above, elongated, adpressed, tubular below ; immersed calices found only cpiite at the bases of the branches. They contain 6 septa, all narrow, and in No. 359 some secondaries. These details seem to indicate that we are dealing with another species, but it is not possible to be certain on the evidence. I can only point out that in other corymbose forms the disappearance of the stem from exposure to the rush of spent waves (not to the breaking waves, under which, in Taliaiti, Aeropora does not live) does not result in a bushy growth, but in the fusion of horizontal branches and reduction of the length of the upright branchlets. Aeropora glochiclados (Brook). (Plate XXXIX, figs. 1-2.) 1893. Brook, p. 104. (No figure.) This species, represented by one specimen, No. 64, differs from A. aculeus Dana in its larger branches and much larger radial corallites, and in that the “ surface and walls are striate and echinulate,” the echinules being fine and pointed. Whether these characters are specific cannot be said without further evidence. The fact that this specimen, in the sizes of the radials, comes nearer to A. aculeus is not sufficient alone. The radial calices are 2 mm. in diameter and 3 mm., rarely 4 mm. long ; walls moderately thick except near the stem, apertures nearly round ; with 12 septa so well developed as to be visible to the naked eye : directives meet deep down, the other primaries nearly meet. Axials hardly 1 mm. exsert, their outside diameter 2 nun., inside 1 mm.; walls spongy, eostulate, 2 cycles of septa well developed. As tins species has never been figured, and A. aculeus not since Dana’s time, I give two photographs on Plate XXXIX. The branchlets below the main branches are much longer than in Dana’s figure of A. aculeus, and so, presumably than in his specimen of A. glochiclados, I should emphasize the striation of the thecal walls, with their fine-pointed echinulae, both of which become coarser on the coenenchyme of the main branches, The thecae “ shorter and sub immersed vi, 3. * 15 214: GREAT BARRIER REEF EXPEDITION below ” according to Brook, on the main branches are completely immersed but for a low rim proximally. Distribution : Brook had only one specimen, from the Indian Ocean. If A. aculeus is synonymous with A. glochiclados, add Philippines, (?) Fiji, (?) and China Sea. Now Great Barrier Reef. Acropora surculosa (Dana). (Plate XXXVIII, figs. 2-5.) 1848. Madrepora surculosa Dana, p. 445, pi. 32, figs. 4 and 5. 1893. M. surculosa Brook, p. 104. 1893. M. recumbens Brook, p. 106, pi. xxvii, fig. f. The three specimens of this species, Nos. 287, 288 and 339 (all with the additional mark W.2), form a very interesting series showing the variation in solidity of the base of corymbose species and the structure and delicacy of the branches and calices induced by exposure to the surf. This series begins with a netted plate ; then one with slits remaining between the branches ; and finally a solid mass, 28 mm. thick at a distance of 5 cm. from the edge, with 3 slits 3 cm. long near the margin, (cf. Crossland, 1928, p. 723, pi. ii.) These effects are sufficiently shown on Plate XXXVIII, but those of the detailed structures of calices, etc., need verbal description. The figures show, however, the great variation in the thicknesses of the branchlets. I have no direct information of the conditions under which these specimens grew. No. 287, the most open form, is from a “ Nigger Head ” off June Reef, while Nos. 288 and 339 are from the outer moats of J une and Ribbon Reefs, all parts of the outer barrier. These more massive forms have the more delicate thecal walls ; in 288 they are very delicate and perforate, in 339 less so, in 287, though still delicate and porous, the lower ones are distinctly thicker. Similarly the development of the septa : in all the axials there are six rather narrow septa, of which the directives are broader ; with, sometimes, representatives of the second cycle, best developed in No. 287, where the directives may meet. In the radials twelve narrow septa are well developed, the lower directives not much broader. In No. 339, the most solid, most radials are quite empty, or with a narrow lower directive, though a few show rudiments of the others. In No. 282 very few radial calices show any traces of septa. Matthai has labelled Nos. 287 and 288 as A. arcuata, apparently regarding this as a synonym of A. surculosa. Brook’s description of this species is overmuch occupied with its horizontal growth form, but there are differences in the shapes of the radials, etc., which separate the species, at least until a more complete comparison is possible. No. 339 Matthai labels A. baeodactyla, a caespitose form differing also in similar details. This specimens corresponds well with A. recumbens. Distribution : From Mergui (Indian Ocean) to Tahaiti. Acropora macrostoma (Brook). 1893. Brook, p. 105, pi. xix, fig. b. Specimens No. 240 “i W moat” and 99 are too young for certain identification, and the group of species to which this belongs needs revision. We know little about the possible changes which occur during later growth. Both these studies can only be properly done on the reef. MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 215 The correspondence with Brook’s description is, however, exact, except that only the youngest coralhtes have walls very thin, though all are spongy, and the growth form of these yoimg examples is purely caespitose, the base being nearly as broad as the area covered by the branches. Distribution : Recorded only from Mauritius, but probably already brought from The Cfreat Barrier Reef of Australia under some other name. Note on Acropora hyacinthus Dana and A. cytherea Dana. 1849. Dana, pi. 32, fig. 2. 1925. Hoffmeister, pi. 13, fig. 3, and pi. 14. 1924. Mayor, pis. 2, 3, 5, 12, 19 and 20. Neither species is present in the collection, though Stephenson and others mention the former on p. 83, and on pi. 24 show very large corals given this name ; and S. Manton has a number of references. Dana’s species differs from the above figures not only (1) in the branchlets of the lower surface but (2) in having thicker and shorter branchlets above, also (3) in the radial calices which are “ not fragile, lip not at all flattened,” and (4) in the axials being short. Stephenson’s (1931) beautiful photograph (pi. 24) of the top of a Paradyce pillar to leeward of Yonge Reef is very suggestive of A. cytherea , apparently showing the branchlets in the little clusters characteristic of that species. In Tahaiti both A. cytherea and the species figured by Hoffmeister and Mayor are common. A. cytherea alone grows to sizes similar to those in Stephenson’s photograph, i.e., usually 1 to 2 metres and more across, the species figured by Hoffmeister and Mayor to a foot or two at most. Hoffmeister describes the young colonies of his species as bushy, but Mayor’s figures as well as his own show that the future formation of an excentric stem is indicated at a very early stage : there is nothing like the small bushy or caespitose forms tentatively included under A. armata here. One of the distinctions of A. cytherea is that the corallum begins as a stout pillar, in which fusion of branches is merely indicated by grooves. (My specimen of this stage from Tahaiti is at Cambridge.) The species differ also constantly in colour, A. cytherea being uniformly lilac pink, A. hyacinthus “ brown purple with pink tips ” (Mayor). Hoffmeister’s species is very near to A. armata, but differs (1) in having much more fusion of the branches, and especially (2) in that the tubulars of the underside are adpressed, and (3) the projecting branches of the stems, with their specially developed tubulars, are not found. Acropora patula (Brook). 1893. Brook, p. Ill ; pi. ix, fig. e. Sample No. 62, without exact locality, corresponds in every way to Brook’s description except that the septa in the radials are more prominent. In the lower radials, where they are well developed both here and in Brook’s specimen, it is the upper directive which is the broader. The species is very near A. latistella, but differs distinctly in the nariform radials which ” remain more or less slightly prominent quite to the bases of the branches ” (Brook). On the main stems immersed corallites, 1 mm. and less across, are rather far apart, on the underside still more widely scattered. Tubulars bearing small buds are conspicuous 216 GREAT BARRIER REEF EXPEDITION on the underside, or on part thereof, and may be 9 mm. long, against 4 mm. in Brook’s specimen, but this specimen, No. 62, is only from the edge of a specimen. I do not find the distinction Brook makes in “ the condition of the surface.” Both are strongly striate and somewhat porous. The Great Barrier Reef specimen of A. latistellci perhaps is more echinulate, but there is no real distinction. A long series might lead to the merging of the species, but on present evidence they are distinct. No. 63 may be this species. Branches much more slender ; stouter, more adpressed radials ; in this more like Brook’s figure. Radials of equal size, meeting. No tubulars on underside. Distribution : Like A. latistella, A. patula is known only from the Great Barrier Reef. Acropora latistella (Brook). 1893. Brook, p. 112, pi. ix, fig. b. 1898. J. S. Gardiner, p. 261. 1925. Hoffmeister, p. 65, pi. 15, figs, la, 16. No. G.B.R. 85 is easily recognized by the naked eye from the characters and distribu- tion of the radial calices alone. Further examination shows complete agreement with Brook’s description, especially since the strongly developed septa are in two cycles in the radials — a rather rare character ; suitably lighted the septa are often visible to the naked eye. Brook says, “ Radial walls of moderate length, but becoming reduced to a crescentic rim near the base of the branchlets.” In most cases the crescent is a complete circle, the inner wall generally well formed, as is quite well shown in Brook’s figure. The pro- jecting labellate radials are found only near the tops of the branches. As their walls are thin the subimmersed calices are very open and conspicuous. Distribution : Great Barrier Reef of Australia, Funafuti and Samoa. Acropora squamosa (Brook). 1893. Brook, p. 120, pi. xx, fig. b. 1918. Yauglian, p. 173, pi. 72, figs. 1-3. Seven fragments broken from the edges of corymbose plates, and showing underneath the short irregular horizontal branchlets the fusion of which forms the plate ; and one larger fragment, possibly the fourth part of the colony. The numbers are P.12, 90, 105, 336, 236 and 285, the larger specimen being No. 107. No. 396 is from A. Reef, Lizard Island. No. 109 is complete though small, 13 cm. across. For relationship to A. sarmentosa see Vaughan, p. 174, and to A. hebes under that species on p. 133. In No. 285 the branches are thinner than usual, and there are no plate-forming branchlets beneath ; but the stems are procumbent, and all calices on the underside are immersed. Higher up branches fuse twice ; they are not in actual contact, but the narrow space between them is bridged by outgrowths. The specimen is thus more like Brook’s than Vaughan’s, and so less like A. sarmentosa. No. 107, the larger piece, and No. 396 correspond well with Brook’s figure. Brook says : “ Axial corallites cylindrical, 2-5 to 3-5 mm. in diameter, wall thick, star very well marked, the directive septa rarely more prominent than the others.” In specimen No. 107, in which none of the more central branches are present, the axial corallites are MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 217 fairly regularly 3 mm. in external diameter, the mean of Brook's measurements, but the thick walls reduce the thecae to 1 nun. internal diameter. The axial septa are distinct but narrow ; usually the primaries alone are present, and even this cycle is sometimes incomplete: but, on the other hand, rudiments of the second cycle are sometimes visible. I have seen broader directives only once in No. 107, but they are usual in No. 396. The septa of the radials are always narrow, sometimes quite rudimentary, and the lower directives not always prominent ; so in this specimen the conditions that Brook found only on the central branches are here present on the outer ones. Certain branches on the central side are remarkable in having the axial calyx filled up with coenenchyme. They are not otherwise deformed or stunted in any way. The same thing is found in some specimens of, e.g., A. variabilis. Distribution : 41 East Indies ” (Dana), Great Barrier Beef of Australia. Acropora hebes (Dana). 1848. Madrepora hebes Dana, p. 468, pi. 35, fig. 5. 1893. Madrepora hebes Brook, p. 128 (no figures). 1918. Acropora hebes Vaughan, p. 174, pis. 73 and 74. 1925. Acropora hebes Hoffmeister, p. 57, pi. 9, figs. 3 a, 3b. Five small specimens correspond to the above descriptions and figures, viz., Z.7, Low Isles moat ; B.M. 278 ; B.M. 294 ; B.M. 288 (the last three have no G.B.B. number or locality) and G.B.R. 294 (B.M. 35) with long lax branches. No. Z.7 (B.M. 50) is especially interesting, as it corresponds very closely with Dana’s figure, and to Vaughan’s photograph of Dana’s type. The fact that it has been labelled A. squamosa by Matthai shows that these two species differ only in (1) mode of growth, (2) the greater regularity of the arrangement of the radial calices, and (3) the absence of immersed calices among the usual exsert ones in A. squamosa. The present specimen appears to be complete : four main stems (one now lost) rise from a common base and the whole is 8 cm. high, being thus, as well as in detail, very like Dana’s type. But his type has more regularly arranged radials and fewer immersed calices among them, thus being more like A. squamosa than is this G.B.B. specimen Z.7. In larger specimens the form of growth is a decisive specific difference. Such a colony as Z.7 might be imagined to grow up into a corymbose form ; but this is no reason for merging branching and corymbose species until evidence for each case has been pro- duced, since it is the rule throughout nature that almost identical young forms differentiate as they grow older. Sample 294 (B.M. 35) differs remarkably from the above in its long lax branches, resembling Vaughan’s pi. 74, fig. 2 of a specimen from Murray Island. Both Vaughan’s figures and Hoffmeister’s account of Mayor’s experiments show that the ends of the branches may be blunt or tapering according to conditions of growth. Vaughan’s photograph of an apical calyx shows only twelve short septa, the descrip- tion stating that the primaries meet deep down. I find that, though all six may meet, usually only four do, and they occasionally meet quite near the surface. For example, in specimen B.M. 294 there may be the twelve narrow septa shown by Vaughan ; or the septa may be much broader at the top ; from three to five primaries may meet quite a short distance down. Distribution : Malacca, Murray Islands, Great Barrier Beef, Fiji, Samoa. 218 GREAT BARRIER REEF EXPEDITION Acropora palifera (Lamarck). 1893. Madrepora palifera Brook, p. 131 (no figure). 1893. M. hispida Brook, p. 133, pi. ix, fig. c. 1907. Isopora hispida Bedot, p. 262, pi. 42, figs. 235-239. 1918. Acropora palifera Vaughan, p. 178, pis. 78 and 79 (all figs.). 1925. A. palifera Hoffmeister, p. 69 (no figs.). The identity of A. palifera with Brook’s A. hispida is shown by Vaughan who discusses other possible synonyms which might be proved by further collecting, but he had a good series of this species and, so far, finds it distinct. Stephenson and others (p. 83, 85 and 88) refer to this species. Dr. and Mrs. Stephenson found it as encrusting sheets yards in extent on the outer ridge of Yonge Reef, in the anchorage coral zone ; and common “ under several growth-forms ” (some of which may be separate species) on most parts of the reef. It is a massive encrusting or lobed form unlike the other species of the genus. “ Also on a reef near Lizard Island.” S. Manton (p. 305) found it also in the moat of Yonge Reef and anchorage zone, her plates xv and xvi showing growths 27 cm. and 54 cm. across. In comparison with these the samples brought home are very small, the larger 12 cm. and 14 cm. high. They are all from parts of the Outer Barrier, and all of the same species, though one (No. 322) is labelled A. plicata by Matthai. This is a solid plate with narrower plates at right angles. It resembles plicata in bearing low ridges, along some of which run lines of tubular calices resembhng apicals ; but they are in contact with one another, have nothing resembling rosettes around them, and all the other calices have decidedly reduced upper walls and thick lower lips, the typical palifera type. There is little doubt that a larger collection would have contained plicata , which both Brook and Vaughan record from the Great Barrier Reef. No. 317 is a cushion-shaped incrustation, 42 mm. thick (and therefore not a young corallum), very closely covered with short tubular thecae. It is labelled by Matthai “ \ Acropora ocellata (Klz.) ” but there are ecological as well as morphological reasons for regarding it as a form of A. palifera. It is unlike any of the numerous figures published. A. ocellata is a stumpy bushy form, and this specimen No. 317 might be a very highly modified surf form. If such forms occur in the Red Sea, Qoseir, where Klunzinger collected, would be an excellent locality for it, but apparently Klunzinger did not find it ; nor have I seen it, or anything resembling palifera at either Ghardaqa or Dongonab. Again, were such a reduction of branches possible under the influence of surf it would surely be found on the Tahaitian barrier edge, where this influence is at its maximum ; but from what I have seen there the utmost reduction would be to a form like plicata, with terminal calices and rosettes round them well traceable, as shown by me (1928) in pis. ii and iii. (Neither A . palifera itself nor any of its related species occur in Tahaiti, though palifera is recorded from Samoa.) I can only conclude that this specimen, No. 317, in spite of its numerous thecae with completely circular walls, among which those with incomplete walls on one side are less conspicuous, is a part of those encrusting bases referred to by Hoffmeister and the Stephensons. In this connection note Hoffmeister’ s remark on the variation in the septa ; in this specimen all the calices are practically empty, only small teeth at the margins perhaps indicating 12 septa. Distribution : Diego Garcia (Indian Ocean) to the Philippines, Rotuma and Samoa. Murray Island and now Great Barrier Reef. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 219 Acropora canalis Quelch. 1886. Madrepora canalis Quelch, p. 150, pi. ix, figs. 2, 2a, 2b. 1893. M. canalis Brook, p. 136. 1927. A. canalis Faustino, p. 275 (who only quotes Brook and copies Quelch’s figures). The species is characterized by a rather loosely arborescent growth, the main branches procumbent bearing upwardly curving branches. These are blunt, with thick-walled axials, 3 to 3-5 mm. in diameter, 1*5 mm. internally, hardly at all exsert. The radials are peculiar, in typical examples the lower lip projects beyond the narrower part containing the septa, is thickened and porous as drawn by Quelch, and this outer part is usually wider than the inner. The septa are well developed, two cycles even in the radials, though the second is usually incomplete. The present collection of eight specimens varies in an interesting way. No. 100, the largest, 17 cm. across, is quite typical. Near it is No. 253, in which the radials have rather thinner lips and septa that often meet an irregular nodule representing a columella. No. 341 is also procumbent, as shown by the more immersed calices of one side ; but above has numerous short branches, only 15 to 20 mm. long, and thinner than in the type : the radials are smaller with thinner walls ; radial septa 6 plus hi number and not meeting ; there is no rudiment of a columella. Tins specimen No. 100 is from the Outer Barrier, Kibbon Reef, inner moat (the preceding presumably from Low Isles), and probably owes its bushy form to rougher water. No. 106 seems to have grown in an upright position. The branches are of normal thickness but more elongated and pointed. A few of the axials appear exsert, but they have radial buds close to the tops. The septa of the radials nearly reached the edge of the lower lip and are 6 plus in number ; but the upper directive is often prominent, and sometimes, deep down, a columella rudiment may be seen. No. 293 (“ June Reef, outer Madrepore field, Madrepore E ”) has four long (5 to 7 cm.) branches with prominent axials, which, under a lens, are seen to bear minute buds. The lower lips of the radials are generally thin, but still a little thicker than the rest of the wall ; the septa are degener- ate, but the upper directive is the largest. No. 455 (E. Moat, Low Isles) is normal, except that the crowded radials and their short lips give the branches a peculiar appearance, almost like a honeycomb. No. 282 (Outer Barrier, Ribbon Reef E. Anchorage) is peculiar in the thinness of its upright branches, and smaller calices. The axials are well exsert, but as in the preceding bear small buds, and their walls are rather thinner than usual. Radial septa well developed, 6 plus in number, upper directives sometimes conspicuous. Lower lip of radials as in preceding specimen. No. 72 is abnormal, consisting of a branch which, having been broken off, continued growth and fused to the broken pieces upon which it fell. This fact may account for some irregularity in the sizes of the radials over a part of the specimen, their regularity elsewhere being a feature of the species. Its main branch is only 15 mm. thick (against 20 mm. in the type), apical branclilets 4 to 5 mm. (against 7 mm.), radials 2 mm. wide and long (not 3 mm.), axials 3-5 mm. and internal diameter 1-5 mm. as typical. Small radial buds appear immediately below the top of the axial calyx. The thin branches of these two last specimens give them a resemblance to A. aspera, but the septa, among other details, show a constant difference. Distribution : Philippines ; and now Great Barrier Reef. 220 GREAT BARRIER REEF EXPEDITION Acropora fruticosa (Brook). 1893. Brook, p. 138, pi. xviii, fig. A. Specimen No. 387 is thus determined by Matthai. It corresponds with the above description exactly, except that the axial calices are narrower, usually 3 mm. instead of from 4 to 5 mm. across, but, as we have seen in the case of A. pacifica, this feature is sometimes variable. Brook says “ Radial calices . . . a little compressed.” This refers to their external shape, but the thickening of their walls makes their openings quite elongated and narrow, as may often be seen in his figure, though perhaps they are more so in this specimen. Distribution : Recorded only from the Great Barrier Reef and Samoa. Acropora spectabilis (Brook). 1893. Brook, p. 141, pi. xviii, fig. b. The five specimens of this species fall into three sets. Three specimens labelled by Matthai are more like the type than the others in having thicker lower and side-walls of the radials : Nos. P.14 and 98 have the walls of the usual thickness, while those of No. 104, in which some of the branches are dead at the top, are only 1 cm. thick, and the radial septa are generally so narrow as to be difficult to see. In the other two the radial septa are much better developed, and a second cycle is visible. In No. 98 the primaries generally meet and the secondaries, though narrow, are complete ; in P.14 the colony seems to have grown in a crack, and is deformed into a vertical plate, and the two cycles are complete, but narrow. Two other specimens, Nos. 204 and 372, have slightly thinner walls to the radials ; and the septa of the radials, except the directives, are narrow, though quite distinct. They are small colonies, and less regular and flat on the upper surface, which may be an effect of youth, but the type specimen probably owes its short and regular branches to a surf-swept habitat. Compare pi. xxiv, fig. 2 of Stephenson and others, which most probably shows this species (not A. gemmifera), but with far longer branches than any brought home, the habitat shown being sheltered from the surf. Distribution ; “ The habitat of the type-specimen is unknown ” (Brook). Now Great Barrier Reef. Acropora gemmifera (Brook). 1893. Brook, p. 142, pi. xxi. Specimen No. 337 (labelled by Matthai “ ? A. gemmifera (Brook) may be A. spectabilis ”) comes from Outer Barrier, Ribbon Reef, outer moat, L.2. It is a fragment with a very solid base, evidently surf- swept. The differences between this species and the preceding are small, but apparently constant. The axials are much narrower, and their twelve septa narrower and irregular, directives often meeting. Outsides of axials and radials striated. In this specimen the septa of radials arc fairly distinct, but are usually made of rows of spines. For “ radial corallites stouter and more proliferous ” I should write “ more irregular in length and thickness.” Distribution : Brook gives six records from the Great Barrier Reef. Also Fiji and Arafura Sea ; not recorded further east. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 221 Acropora humilis (Dana.) 1848. Dana, p. 483. 1849. Dana, pi. 31, fig. 4 ; pi. 41, fig. 4. 1893. Brook, p. 145. 1925. A. samoensis Hoffmeister, p. 60, pi. 11, fig. 4. This species lias a great resemblance to A. digilifera , in its growth form ; thickness, length, and blnntness of branches, and proportions of the axials, so that it is very easy to confuse them. They differ in the form of the radials, absence of immersed calices except right at the bottoms of the branches, and especially in the well developed septa, which in A. humilis, are so well formed, so as to be visible even to the naked eye. Dana’s drawing of them shows the complete bilateral symmetry of the oval apertures due to the nearly equal directives coming close together centrally : but it is peculiar in showing 8 primaries, a mistake which is probably due to the fact that, in some cases, one of the secondaries may be enlarged, though usually they are narrow, and the cycle incomplete. Dana also states “ star scarcely distinct,” which is at variance with his drawing ; but perhaps he means that the radial symmetry is quite lost. The axials have 12 septa, of which 6 almost meet, and the secondaries too are broader than usual. They are thin and somewhat irregular, may be turned at a sharp angle at their inner edges, and sometimes bear an appendage that stands, as noted by Brook, at an angle with the plate. Directives are distinguishable, but not usually prominent. The two specimens of tins collection are larger than Dana’s type, which is obviously young, and have branches up to 8 cm. long, and rather more tapering. No. 73 (no locality, presumably Low Isles) is complete but for one lost branch. No. P.10 is a forked branch with three short branchlets, and differs from both the type and No. 73 in having shorter and thicker radials, and more regular septa in the axials. Both show occasional arrange- ment of the radials in rows, as in Hoffmeister’s photograph. Distribution : Fiji, ? Samoa (Hoffm.), Great Barrier Beef. Acropora bruggemanni (Brook). 1893. Brook, p. 145, pi. xxiv (with synonymy). Three small coralla, of which the largest (No. 256) is 11 cm. high, and therefore very much smaller than Brook’s figured example, and seems to be the whole of a young colony. Nos. 247 and 257 are about the same height, but consist of only a few branches each. They are all labelled by Matthai. This is an easily recognized form, and is mentioned by Stephenson and others but not by S. Manton. It is therefore rare or absent from Low Isles, and Dr. and Mrs. Stephenson write of Three Isles : “ Only one common coral species wTas noted as a form not familiar to us at Low Isles. This was an important form ecologically — a species of Acropora (A. bruggemanni ) occurring in quantity on part of the flat adjacent to the anchor- age, and making platforms like those of A. hebes at Low Isles (the latter species was also present).” This, like the occurrence here of three alcyonarians, is another of those unex- plained differences between reef faunas already noted. Three Isles is a long way north of Low Isles, but it is similar in structure and similarly situated on the Barrier platform. All these specimens belong to the typical blunt-ended form, but No. 256 has springing from its base, three short branches (the longest 3 cm.) which are tapering and pointed. 222 GREAT BARRIER REEF EXPEDITION Brook does not mention the irregular broadenings of many branches ; the species clearly leads on to the Isopora group, and a branch in this specimen 35 mm. broad and 38 cm. high carries on the resemblance, especially as this is one of those branches in which an apical theca is not distinguishable. The septa differ from Brook’s description in that (1) the primaries are often narrow, and so do not differ greatly from the secondaries, and (2) the directives may be broader in both axials and radials. Their edges are curiously wrinkled. Distribution : Not recorded since Brook wrote, and only from Singapore besides a number of localities on or about the Great Barrier Beef. Acropora variabilis var. pachyclados Klz. (Plate XXXVIII, figs. 1, 6.) 1879. Klunzinger, p. 17, pi. ii, figs. 1 and 5 ; pi. v, figs, la, 16, 3 ; pi. ix, fig. 14. 1893. Brook, p. 161. 1906. Marenzeller, p. 49, pi. 15, figs. 40 to 44. 1918. Vaughan, p. 181, pi. 80. Three specimens, Nos. 343, 344 and 348, correspond closely with Klunzinger’s description of his thick-stemmed variety ; of his figures pi. ii, fig. 5, pi. v, fig. lb, and pi. ix, fig. 14 afford good illustrations, the last giving the shapes of the calices. Of Maren- zeller’s figures pi. 15, figs. 41a and 43 are nearest to these specimens. Vaughan’s figures are of a thin-stemmed form. The significance of the name variabilis is shown in these samples by the existence, among the thick stems, of some much thinner. In No. 344 the stems are mostly from 2 to 2-5 cm. thick, but counting only those springing from the base , there are two averaging only 1-5 cm. In No. 348 the thicker stems are only about 1-5 cm. thick, and four are only 1-0 cm. In this specimen there is a tendency for the larger radials to fall into rows, as in Marenzeller’s figure 43a. This is a peculiarity, which reaches an extreme in No. 343, where, not only are the rows quite definite, but there is a fusion of the bases of the radials, (sometimes almost their whole lengths are involved), so that they form almost plate- like structures, bearing also small buds and half immersed calices. This arrangement is curious and unique, deserving illustration, and perhaps even a varietal name, but unfortunately only three detached branches are present. In Nos. 344 and 348, especially in the latter, some of the tubular radials taper towards their tops ; in the latter many of the adpressed tubulars low down on the stems are almost bulb-shaped. The axials are up to 6 mm. thick, with rounded, smooth, finely-granulated walls, and their internal diameter 1-0 mm. and less. In No. 348 all but two of the main axials are filled in with coenenchymal matter, leaving only a trace of the cavity, or none at all. In No. 344 the cavities are usually 0-75 mm. across, or may be oval, 1-0 mm. x 0-75 mm. The septa have not been described. In the axials two cycles are complete, directives broader, nearly, but not actually, meeting ; laterals well developed, in the two axial calices available in No. 348 the septa are broad but thin, and rather irregular ; in No. 344 some calices have narrow septa. In the tubular radials again there are two cycles of septa, the second not always complete. In No. 343 the lateral septa are well developed, but in the others this is rare, laterals being usually narrow or even rudimentary. The immersed calices have also MADREP0RARL4, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 223 two cycles of septa, all very narrow (no directives) so that they look like open holes : these are very conspicuous in No. 343, but less so in the other two, in which they are surrounded by a low, ring-like wall. The characteristic buds on the larger radials figured by Klunzinger are not always present in No. 348. Distribution : Red Sea : Philippines, to Tongatabu, blit not further east. Vaughan gives Samoa, but Hoffmeister does not record it. Acropora exilis (Brook). (Plate XXXIX, figs 3, 4.) 1893. Brook, p. 172, pi. x, figs, c, d. The species is cpiite abundant at Low Isles since S. Manton referred to it on pp. 287, 295 and 296, and on pis. iv, vii and viii, and shows numerous examples on pi. xi. It is common at the seaward ends of all three traverses ; on Traverse I ££ The most abundant species as measured by the number of separate colonies. It forms bush-like growths 3 to 12 inches across.” On Traverse II " particularly abundant at the seaward edge of the reef.” This local abundance is striking, considering that the species has not been recorded any- where else since Saville Kent's time, and is a good instance of the way in which coral faunas vary from one area to another. Of four specimens labelled by Matthai, Nos. 69, 83 and 199 (" Tl outside ”) are typical, Nos. 69 and 83 corresponding to Brook's figures of bushy and long-stemmed (“ virgate ”) forms, while No. 199 is intermediate. These two long-stemmed forms show that this is one of the species which, like A. elseyi and others, may grow up from some depth, dying below as they extend upwards in a crowded mass of branches. It would be interesting to know whether the lower dead parts differ in any way from the living bushes above. The bushy specimen has made a fresh start, though growing on a broken and dead fragment of its own species. The fourth specimen, No. 45, bears little resemblance to the others. The branchlets consist of stout, tubular axials bearing a few small short buds, but are naked for the distal 3 or 4 millimetres. On the main branches the radials are mostly long tubules pressed to the stem except near the ends of the branches, where their upper ends are free. (Plate XXXIX, fig. 3.) The septa are fewer, 6 in number plus 1 to 4 or 5 secondaries. Never- theless the identity is clarified by the basal and some inner branches of the bushy specimen, the only one with basal branches in which similar conditions are found, though not including the long adpressed radials. However, as intermediates to these radials occur on normal branches we may consider this curiously abnormal form as belonging to this species. It is unfortunate, however, that the whole colony was not brought home. Distribution : Brook records it from the Great Barrier Reef, the China and Arafura seas. Though a very beautiful and easily recognized species, it has not been recorded again until now. Acropora elseyi (Brook). 1893. Brook, p. 172, pi. xi, figs, e and F. No. G.B.R. 307 (June Reef Anchorage, coral zone, p.) is one branch with three divisions. No. B.M. 33 is a small but complete colony growing upon broken pieces, probably of the 224 GREAT BARRIER REEF EXPEDITION same stock. As Brook’s fig. e also suggests, this is one of those species of which specimens die off below as growth proceeds upwards. These specimens are about half the size of those figured by Brook, and No. 33 seems to be young, as the upper raclials nearly all have clean-cut edges ; but in No. 307 they are nearly all thick and rounded. Below the radial calices thicken and finally become “ verruciform,” or rather conical, with very small apertures which sometimes disappear altogether. In these peculiar basal corallites, as in the septa and the densely echinulate coenenchyme, the species differs from A. polymorpha, to which species there is a strong naked-eye resemblance. A few other basal corallites are immersed, wide open (1 mm.), with slightly projecting walls. There are six well-developed axial septa. Distribution : N. Australia and Torres Straits. Saville Kent collected it from Rocky Island, the position of which I am unable to find. Acropora rosaria (Dana). (Plate XL, figs. 1, 3 to 4.) The species, usually rare, is referred to by S. Manton on pp. 287, 295, 305, and plates vii and xiv : the last figure shows a single small specimen on the crest of Yonge Reef. On Traverses I and II it is one of the rarer species in the deeper parts. The series of 7 specimens: — No. 310, Ribbon Reef; 334, June Reef anchorage; 368, Ribbon Reef; 108 and 77 no locality ; 207 and 216, “ Tl outside” ; and 276, June Reef “ Inner moat, 0 2”; (the first three and the fifth and sixth also bear the sign 9). All but Nos. 206 and 334 have been labelled rosaria by Matthai, but No. 108 is labelled “ A. spicifera (Dana) may be A. rosaria (Dana),” yet seems quite clearly to belong to the latter species. These 7 specimens fall into three well-marked forms. That the species is variable is noted by Brook, who describes three forms, the first divided into two varieties, only the last of which, forma dumosa, he illustrates : it is not found in this collection. I there- fore cannot identify them here, but Nos. 368 and 108 may be his forma rosaria var. diffusa ; and Nos. 206, 216 and 276 may be his forma pygmaea, of which S. Kent brought four specimens from Low Woody Island. None here are exactly like Dana’s type, a photo- graph of which is given by Vaughan, in pi. 82, fig. 2. I therefore describe the series afresh. Form I. — (Plate XL, fig. 3.) Nos. 310 and 334 are alike ; both are frotn the outer barrier, Ribbon and June Reefs. They have the typical growth form with the ends of twigs standing out nearly at right angles to the main stems. Axial septa 6 plus 6, primaries nearly meeting, secondaries comparatively broad ; directives generally distinguishable. Radial septa also 6 plus 6, all narrow, directives the broader, the upper one usually best developed. Septa of immersed calices have both series of septa, also narrow ; but being thicker, they are more distinct. These immersed calices are frequent more than half-way up the branches on the fronts of the specimens, nearly to the tops on their backs. In these characters the specimens are typical, but they have some additional characters the most striking difference being that the axial calices are extraordinarily broad — 4 to 5 mm. (generally the latter) in No. 310, 3-5 mm. in No. 334. As in other specimens (the exceptions are noted below) the cavities of the axial calices are only 1-5 mm. wide. MADREPORARIA, HYDROCOBALLTNAE, HELIOPORA AXD TUBIPORA— CROSS LAND 225 The walls of the radials are also thick; many are short tubules, 2 mm. long, pressed to the stem. Coenenchyme porous, with echinulae in rows, thecae finely striate, but scarcely c-ostulate. In both specimens the backs have thinner, shorter branches with narrower axials, and little but immersed calices on the main stems. Form II: — Nos. 108, 368 and 77 (Plate XL, fig. 1.) are complete colonies, bushy growths 10 and 17 cm. high, with encrusting bases overgrowing dead coral — hi the case of No. 108 probably of the same species. In spite of their irregular bushy forms one side is flatter than the other, with thinner branches and axials ; in No. 108 and on the back of 77 these are cpiite long and slender, some of the branchlets being long, tubular axials bearing only a few small buds ; but none are quite naked as in Brook’s forma dumosa. In these specimens the normal axials are 3 mm. thick, with cavity 1-25 mm. across. Radials are also thinner-walled than in the preceding, and many are short, adpressed tubules. Immersed radials are numerous at a point more than half-way up on both sides of the colonies. Axial septa not always 6 plus 6. In twelve main axials examined there are four with this formula, one with 6 plus 5, two with 6 plus 3 and one with 6 plus 2, the remaining one having 10 septa rather irregularly arranged. The radial septa are arranged as before. Thecal walls porous, but thecae not striate ; coenenchyme with echinulations in lines, but compact. Form III : — Specimen No. 216 (Plate XL, fig. 4) and especially the fragments No. 276 and 207 differ in that most of the radials are 4 mm. long, and pressed to the stems. Axials narrow, 3 mm. externally, and only 1‘0 mm. internally, as are the radials. In No. 216 the lower branchlets have, especially on some dead branches at the base, small radials and elongated axials which have an appearance quite unlike those of this species. Many of the adpressed tubular radials have broad septa like the axials, as though the distinction between axials and radials were not developed. In No. 207 the septa of the second cycle are few, often absent in the radials, in which the directives are often prominent. These three specimens have faintly striate thecal walls, prominent echinulae on ridges over the coenenchyme, which is sometimes dense, sometimes reticulate. A. murrayensis Vaughan is only distinguished from this species by a denser coenenchyme, involving thecal walls without costulae, and, in view of the above, this seems hardly enough for a specific distinction. Distribution : Fiji and Great Barrier Beef. Acropora cancellata (Brook). (Plate XLI, figs. 3 and 4.) 1893. Brook, p. 166, pi. xxxii, fig. c. A fine specimen of this species was dredged from Stn. XXVII. It is a contrast to the small piece figured by Brook, the only known specimen, which was dredged over 90 years ago from 15 fms. The correspondence with Brook’s description is exact, except that, being fully grown, the branches at the base are much thicker ; but I give figures of the complete specimen, and details. The density of the corallum is noticeable as soon as it is taken in the hand. This applies to the walls of the tubular corallites, as in A. clavigera. The coral, though 226 GREAT BARRIER REEF EXPEDITION growing from an encrusting base, differs on its two sides as does Brook’s specimen, and therefore may be supposed to have grown out horizontally, or nearly so. Distribution : Louisiade Archipelago, and now Great Barrier Reef. Acropora clavigera (Brook). (Plate XL, fig. 2 ; Plate XLII, fig. 3.) 1893. Brook, p. 183, pi. ix, figs, a, a'. A fragment dredged at Stn. XXV differs somewhat in growth form from Brook’s specimen. No branch is quite so like a rhizome of Caulerpa racemosa as is Brook’s fig. a', and the base does not form a network though the branches are flattened. Different parts of the colony naturally differ in these respects. The Expedition piece is from the edge of a growing corallum. The correspondence with Brook’s figure is otherwise complete ; but I should emphasize the inconspicuousness of the buds on the sides of the tubular calices due to their being small and “ liemicotyloid,” and the thickness and solidity of the walls of the tubulars even close to their ends — a very rare thing in the genus. There is a ring of about 16 costal ridges surrounding the apical orifices ; the 6 primary septa are thin, but not narrow ; of the directives one or both maybe conspicuous sometimes neither. The secondaries are short and distinct, and may be two only is complete. As this is only the second fragment to be collected, and Brook’s figures are obscure, I illustrate it again on Plates XL, XLII. The locality of the original specimen is not known. These last two species are described by Brook, the one under the subgenus Conocyathus the other under Rhabdocyathus. As a matter of fact the species are strikingly similar, the main difference being the greater prominence of the smaller radials and number of immersed calices in A. cancellata. Indeed the separation of the two species is probably only provisional until more specimens can be compared. Acropora prolixa (Verrill). 1868. Madrepora prolixa Verrill, p. 22. 1902a. Acropora prolixa Verrill, p. 237, pi. 36, figs. 3, 3a ; pi. 36 a, figs. 3, 3a ; pi. 36 f, fig. 14. 1925. Acropora prolixa Hoffmeister, p. 65, pi. 16, figs, la-c (with synonymy). Hoffmeister’s is the only figure of a complete colony, and should be supplemented by Verrill’ s of small branches. The single branch dredged from Station XXVII is more loosely growing, and, as one side bears fewer and less developed branchlets it may have been more or less procumbent. It is 11 cm. long, and 8 mm. x 12 mm. at the base (which is irregularly elliptical, as the break passes through the thick bases of a pair of branchlets). A marked peculiarity of this specimen is that its echinulations and the ridging of most of the main stem (there are no ridges on the thecae) are more conspicuous than on any other species I have seen — far more so than Verrill and Hoffmeister describe. In places the large, closely-placed granules are seen to be thicker at the top than below ; the striations are well separated by deep groves. The explanation is found in the section of a branch ; the main stem is nearly solid, but that of a smaller branch is built of several solid rings surrounding the imperforate walls of the axial calyx, the rings connected by stout bars ; and outside all are the echinulate plates. Evidently the stalked granules and deeply separated plates are only a stage in the growth of the coenenchyme. The septa are as described by Hoffmeister, but the whole of the second cycle is not MADE E PORAR IA, HYDROCORALLTNAE, HELIOPORA AND TUBIPORA— CROSSLAND 227 often present. In some cases tlie upper directive is mucli the larger than the other. I agree with Hoffmeister that this species is distinct from M. echinata Dana and the other species listed by him. It is near to M. procumbens Brook, and just possibly identical. Distribution : Samoa, Siilu Sea, now Great Barrier Beef. (Yerrill’s “ Ousima ” cannot be found in the largest atlases.) Acropora jeulini sp. n. (Plate XLII, figs. 1-2, 4-5.) Specimen Xo. 330 (no locality) is a horizontal growth, of three main branches, 2, 2-5 and 3 cm. thick, joined by a base 5x3 cm. in section. The ends of two branches are dead and overgrown, but they widen to 3-5 and 3 cm. where stumpy side branches are given off. The underside bears stumpy branches, more or less horizontal, as shown on Plate XLII, covered with somewhat conical calices, distally tubular, proximal ly mere warts without openings ; a few calices are immersed with projecting ring, generally thicker and higher on one side. The axial calices of the proximal branches are hardly distinguishable. The upper surface bears short conical branchlets, two of which are 2 cm. long and wide, and others range from a mere central tubular calyx with a rosette of radial s round it to branchlets 1 cm. long x 1 cm. thick. Axials thick-walled, walls very rounded, spongy without costulae, 3-75 mm. thick ; cup 1 mm. across, very little exsert, and bearing small buds near their margins. Septa 12, of which the primaries join an irregular columella-like body. In one case the cup is 1 -6 mm. wide, but this is a giant calyx, having 16 septa all of the same size, joining a rough columella. The radials of the branchlets are all tubular ; inner walls thick, the outer very thick and rounded ; apertures round, facing upwards, and therefore oblique. Septa in two cycles, the two lower laterals nearly as broad as the directive between them, the upper directives and upper laterals being much shorter. Walls spongy, only the lower third finely striated. The flattened surface of the main branches is crowded with similar calices, many facing in all directions, but more often distally : in any case they face so far outwards that the openings are conspicuous in a view from above (Plate XLII, fig. 5). Some of the larger, though not cylindrical calices, bear small buds. Among these are completely immersed calices, in which the 12 septa are all equally narrow. Some of the calices become closed up, partly by thickening of the septa, but also by ingrowth of the walls. This is particularly frequent on the lower surface where, proximally, many become mere warts with no trace of an opening. The coenenchyme is reticulate in places, but generally striate and echinulate, with holes in between the striate. The echinulae are flat-topped and bear minute spines, visible under a lens but best seen by binocular. On the proximal part of the main branches these are very numerous, so that their flat tops fit together into a mosaic. This species resembles A. decipiens (the specimen shown by Vaughan (1918) on pi. 67*) A. smithi (Brook) and A. plicata, but is at once distinguished by its thick- walled * Crossland in his typescript did not indicate what figure of Vaughan’s he had in mind. — [A.K.T.]. 228 GREAT BARRIER REEF EXPEDITION and never gutter-shaped radials, besides other structures. It might be a surf-flood modifi- cation of some form like A. secaloides Yerrill or A. abrotanoides (Brook), but this is not likely. The effect of the surf-flood in the Astreans certainly, and in Acropora, etc., so far as my observations go, is not to thicken or shorten the calices (rather the reverse), but to shorten branches and solidify the basal plate only. (Crossland, 1928, p. 723, pi. ii. For Astreans see Crossland, 1931, e.g., pi. v, fig. 9,f and its explanation.) I have named the species after M. Jeulin, who worked as an amateur of corals at Ghardaqa and in the Gulf of Suez. Acropora brool'i sp. n. (Plate XLI, fig. 2; Plate XLV, fig. 3; Plate XL VI, fig. 1.) Specimen No. 290 (? June Reef, outer moat) is a fragment of a corymbose plate, measuring 13 cm. radially and 10 cm. across. Viewed from the underside the branches are flattened and fused, leaving small holes proximally, elongated slits distally, as shown on Plate XLI. The surface is largely smooth, but for scattered warts and immersed calices, but distally completely adpressed tubulars abound. The sides of the branches, however, bear long tubulars 2 mm. thick, which proximally form triangular branchlets, often fusing to the adjacent branches. The upper side bears branchlets 2-5 to 3 cm. long and about 0-5 cm. thick, the proximal nearly vertical, the distal curved upwards. These are placed further apart then usual, their apices 1-5 to 3 cm. distant ; probably in corres- pondence with this the immersed calices of the plate are very numerous, and nariform radials continue on to the plate in lines from the bases of the outer branchlets. The axials are long, 4 mm. x 2 mm., but lower down the branchlets are others 8 mm. long bearing a few small buds, or becoming aborted twigs. They are 2 mm. thick, cavity 1 mm., wall strongly striate and porous. Axials with 6 narrow septa, and small representatives of the second cycle, sometimes complete. Directives sometimes smaller. Radials nariform or compressed, tubular, quickly attaining their full size ; projecting at an angle of 45° to the stem ; walls thin. As they are 4 mm. long they are conspicuous ; width 0-75 mm., shape of opening like a Canadian canoe, and with thin lips ; the opening does not reach the stem, but sometimes is close to it ; in the larger and lower radials it is separated by a variable distance increasing to nearly the length of the calyx in the the tubulars, in which the opening is oval. Walls, like those of the axials, very distinctly striate, the striae continuing over the lips. About half-way down the branches the radials become nariform, with much wider oval openings, and finally become round, subimmersed and immersed. Septa of the radials reduced to the outer directive, which is, at its best, a thin line along the bottom of the “ canoe ” ; and even this is often difficult to distinguish. Tubulars with buds have small laterals as well as narrow directives. The large, immersed calices of the plates have slightly raised walls ; some, near the bases of branches derived from the short nariforms of the lower parts of the branches, are only subimmersed. They are quite open, without septa or with the slightest rudiments. Coenenchyme reticular, echinulate, the prominent striae of the thecae not extending beyond their bases. t Crossland had written “ pi. ix,” which shows Lobophyllia. It seems likely that he meant “ figure 9,” which shows a Leptastrea of the barrier edge series. — [A.K.T.] MADREPORARIA, HYRROC'ORALLIXAE, HELTOPORA AND TUBIPORA— CROSSLAND 229 Acropora lutkeni sp. n. (Plate XLI, fig. 1 ; Plate XLYI, fig. 2.) Specimen No. 271. from the Outer Barrier, June Reef, Nigger Head, is a complete colony, consisting of a solid conical base 11 cm. high, 6 cm. wide at the bottom, with a more or less flat top about 20 cm. across. This bears a number of simple upright branchess about 3 cm. thick, regularly tapering, bearing on their lower halves numerous branchlet, up to 1 cm. long and thick, but mostly much smaller. At one side of the base are branches 4 cm. long, 1 to 1-5 cm. thick, obviously homologous with the branchlets of the main stems. Where the branches come near together they are flattened, so that sections are oblong or triangular. The striking character of the species is the extreme compression of the more prominent radials, 3 to 4 mm. long, the cavities of which are usually deep slits, opening outwards distally and not quite reaching the stems proximally. In other words the radials have greatly thickened side-walls, but their ends are not, as in many species, specially thick. These openings, and the upper surfaces of the calices, are nearly at right angles to the stems, but the lower walls slope upwards slightly. Among these are shorter calices with oval openings, and immersed with ring-like walls. The longer radials bear small buds, and, about 3 cm. down the stems, these buds enlarge to form rosettes, the beginnings of the branchlets of the lower parts of the stems ; but though these radials have now become axials they retain the slit-like shape of their cups. Even the axials of the main branches are oval. The immersed calices with ring-walls are found up to 1 cm. from the tips of the branches ; but lower down they become numerous, occupying most of the space between the branchlets. As then 12 septa are all very narrow their calices are open and conspicuous. They are accompanied by adpressed tubular radials, with thick outer walls and small (0-5 nun.) openings, upwardly directed and close to the stem. The axials of the main stems are 3-5 mm. thick, their cavities 1 mm. longer in diameter ; they have 12 septa, all short, but in most cases accident or abnormal growth make them difficult to see. In the radials only the directives are conspicuous, the lower as a long ridge in the bottom of the trough, the upper as a wedge-shaped tooth on the proximal wall ; laterals are present, but rudimentary. The walls of the thecae and the coenenchyme are finely and closely echinulate, not striate, echinules long and pointed. I name this striking species after the late Professor Liitken, in recognition of his labours in collecting and arranging the corals of the University Museum of Kobenhavn. Prof. Liitken did not publish Iris results, but I hope to give some account of his work later. Acropora olteri sp. n. (Plate XLIII, figs. 1-2 ; Plate XLIY, figs. 1-2.) The three specimens Nos. 306 T2, 338 T2, and 60 (the two former from the outer barrier, Ribbon Reef, inner moat ; and June Reef, outer moat respectively) differ greatly in appearance and largely in structure, but undoubtedly belong to the same species. The growth form is corymbose, with arching upright branches. But while No. 60 is decidedly lax, with no fusions of branches beneath, and only a trace of flattening, the upper branches somewhat irregular, up to 11 cm. long, Nos. 306 and 338 have flattenings vi, 3. 16 230 GREAT BARRIER REEE EXPEDITION and fusions beneath, the former with regular branches 7 cm. long, of which the radials are all approximately the same length, the latter with stout branches, made irregular by numerous long tubulars bearing buds and aborted twigs — an effect which is even more marked in No. 60. These differences are so curious that all three specimens are photo- graphed. These three specimens are the syntypes of the species.* All are alike in having thick, axial calices, walls spongy above, inconspicuous striae on the sides, and moderately wide cavities. External diameters are 3 mm. in No. 60, 3-5 mm. in the others ; internal diameter 1 mm. in all. Septa number 6 plus 6, thin but straight, narrow above except the directives, but broad below. Axials exsert about 2 mm., but may have minute buds within this distance of the apex. The projecting radials are of two kinds, (1) shorter, nariform with oval or nearly round openings, inner wall always distinct, though short, and (2) tubular, with straight or nearly straight-cut ends, the longer of which bear buds. These larger radials tend to fall into rows, straight or somewhat spiral. Lower down the nariforms become adpressed the outer lip thickening and the opening becoming round, the whole becoming kke short adpressed tubulars. Among these are numerous immersed and subimmersed calices, some of which are to be found 1 or 2 cm. from the tips of the branches. In the upper radials only a narrow, lower directive is visible, but, lower down the other septa appear and are often well developed, some of the second cycle being present. The long, tubular radials, with others which are shorter and bear small buds, are a characteristic of several species, e.g., A. abrotanoides ; but the combination of these with the adpressed tubulars and sub- and fully-immersed calices of the lower parts of the stems are a special feature of this species (Plate XLIII, fig. 2). In No. 306 tubulars occur, but are not prominent, and rarely bear buds ; and the lateral twigs are short and stumpy. Exactly the reverse is true of No. 60, in which long tubulars bearing buds are conspicuous, and the lateral twigs are long and thin, often hardly thicker than the axial calyx itself. In No. 338 they are often prominent, but lateral twigs are short. The coenenchyme is finely striate and echinulate, but more coarsely than the thecal walls. The species is near to A. abrotanoides , but its corymbose form is not the only difference. It is not proved that caespitose and corymbose forms occur in the same species when adult : the form of branching of the specimens of A. abrotanoides in this collection and Vaughan’s figures (1918, pi. 68, figs 1, 1 a and 2) make it especially unlikely in this case. , Acropora laevis sp. n. (Plate XLV, figs. 1-2.) Nine specimens, though widely differing in detail among themselves and all differing from A. pulchra, agree in a general smoothness to the naked eye, which is due to (1) the smoothly rounded and thickened edges of the calices, (2) the fine and uniform granulation of the coenenchyme, (3) having well developed stars of septa in the radials, and (4) the conical shape of the half-immersed radials of the lower halves and backs of the branches. Branching is lax, side branches at angles of 30° to 45° to the main stems, and varying * Crossland wrote : “ All three constitute the type.” But in accordance with the recommendations on International Rules of Nomenclature, No. 338 is hereby designated holotype, and Nos. 306 and 60 paratypes. — [A.K.T.] MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 231 in thickness from 12 mm. to 8 mm. basally. Radial s well separated (No. 79) or crowded (No. 93), usually tubular, and sometimes at right angles to the stem, or inclined (compare figs 1 and 2 on Plate XLY). There are also conical and immersed calices high up the stems, but the latter reach their characteristic form 5 cm. or so from the tops, and are the only kind for the lower halves of the stems (Plate XLV, fig. 1 ). The walls of the longer calices are ridged, if at all, only near their bases, and the upper parts are slightly thicker and finely granular. The lower directive is usually the largest septum, but is not so prominent as to prevent the septa making a star. For other characters I refer to the plates, but Nos. 33, 212 and 217 deserve special mention for their large, thick-walled calices, of which the largest are remarkable for their narrow bases : some might be called club-shaped. Holotype : B.M. Peg. No. 1934. 5. 14. 349. G.B.R.E. Sample No. 79. [Family Poritidae.] Genus Gonioporci. S. Manton mentions (p. 289 and graph xi) that species of the genus occur in minute colonies on the muddy rocks of the seaward slope of Traverse I. On pi. xvi she shows a single, small specimen on the inner part of Yonge Beef. Stephenson and others give four references (pp. 67, 86, 89 and 90) listing it as characteristic of seaward slopes and the anchorage at Low Isles, of Yonge Beef, the reef patch at Lizard Island and Batt Beef. In many cases I cannot make out the mode of growth by layer over layer which Bernard so elaborately describes as characteristic of the genus (indeed he implies that it is invariable). Such layer-over-layer formations are found in any genus when a colony is successful in resisting periodical partial burials in mud or sand ; and it is particularly common in this genus, which, as noted also by S. Manton, often inhabits muddy areas. Gardiner (1939, p. 242) also finds that Brook’s “ description is inapt ; for the central part, perhaps inches thick, has no epithecal plates one over another, but is a mass of trabeculae, as described above.’' He is speaking here of G. stokesi, but the same thing is true of three of the species about to be described, and in G. columna, which is common at Gliardaqa, the tall columns, say two feet long, are continuous from top to bottom. Gardiner continues, “Clearly there is no true ‘epitheca,’ based on a ‘primitive epithecal cup’ as Bernard expressed it (p. 18), but a later secondary formation,” a protective adaptation to the special surroundings in which many of the species live. The genus occurs from the Bed Sea to the Pacific. It is recorded as far east as Samoa by Brook, but the record is doubtful ; and Hoffmeister does not describe any species from Mayer’s collection. J. J. Lister, however, found three species in Tongatabu. As my identifications of these specimens with Brook’s 12 numbers from the Great Barrier Beef are all marked “ probably ” I omit them. Goniopora tenuidens Q. 1918. Vaughan, p. 186, pi. 84, figs. 1, 2. No. 254 is the greater part of a fair-sized colony, 12 cm. across x 9 cm. high, smooth and originally hemispherical. The mode of growth is not layer over layer, but it is a thick (max. 4 cm.) sheet folded to a horseshoe section. A delicate epitheca covers part 232 GREAT BARRIER REEF EXPEDITION of the base. The specimen corresponds well with Vaughan’s variety 2 on p. 187 over most of the surface, but there is rather a sudden change near the base, illustrating Brook’s quaint remark that all Gonioporas are Rhodaraeas near their bases : the calices change from being more lightly built than Vaughan’s fig. 1 to the shallow, thick-walled, solidly built calices with huge pali of his fig. 2. Distribution : Great Barrier Reef ; Amboina ; Philippines. Goniopora lobata M. B. and H., Bedot. (Plate XLVII, figs. 1-3.) 1907. Bedot, p. 267, pis. 43 and 44, 3 figs. Specimen P.37 is part of a hemispherical colony 6-5 cm. across ; No. 57 is a fragment of a large colony ; and the third, dredged from Stn. XVI, is complete, but only a flat crust measuring 4 x 2-5 cm. These three specimens differ greatly in structure. No. P.37 is the most like Bedot’s typical examples, which he says are absolutely like those of Milne Edwards and Haime. Bedot’s description is excellent, but the looser structure of the columella in the present specimen enables me to describe an interesting point in its morphology (Plate XLVII, fig. 2). It is large, occupying most of the bottom of the calyx but for holes round the edge represent- ing the loculi, and is flat except a small area in the middle, forming a floor. It has two origins, the central part being made of trabeculae from the edges of the septa in the usual way, this part being about one-third of the diameter of the calyx, and bearing the charac- teristic upright spinules. Outside this and below the edges of the septa, the columella is formed of distinct synapticula, standing more or less at right angles to the septa. Being at a slightly lower level they appear only occasionally in the published photographs, and also are often obscured by greater development of the upright columellar teeth. The specimen dredged from Stn. XVI is very different, having shallow, flat calices divided by flat-topped walls 1-2 mm. thick. The columellar teeth become, in some cases, indistinguishable from pali, plate like or double : the columella is narrower, but its parts distinct. This specimen shows the typical fusions of the septa described by Bernard (1903, p. 21) in which tridents alternate with simple primaries, with unusual distinctness. Sample No. 57 I identify with some doubt, as it is only a fragment of a large colony. It has exceedingly delicate walls, which have been much rubbed, and the large, flat columella is obscured by fusions, but the columellar septal teeth are essentially the same. On the broken edges of these specimens I can see delicate tabulae in P.37 similar to those in Bedot’s figure, but not everywhere in the section. In No. 57 I can see none at all, and of the dredged specimen no section is available. The synonymy of this genus is in a hopeless condition. Bernard’s obsession with growth forms naturally led himto abandon attempts at synonymy, and spoils his descrip- tions. Consequently we have sound descriptions, such as those by Bedot and Vaughan, for very few species. Bedot says the Paris specimens of G. lobata and G. savignyi are not distin- guishable, but Savigny’s drawing and I find in such cases as I know that his drawings are very accurate) represents another (species, so that it seems as if there had been some con- fusion of labels. Klunzinger’s G. planulata is, apparently, a stunted specimen of Dana’s G. columna, such as one would expect to get at Qoseir. In sheltered water at Ghardaqa this species is found in large convex masses, made up of closely fitting columns. This is quite MADREPORABIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 233 different to the present species, and is not at all like Bernard’s pi. xiii, fig. 12, which he calls (p. 100) Red Sea (6) 1, — G. lobata. In G. columna there is no “ expanding column consisting of a thick sheaf of lamellae ” — there are no lamellae. Distribution : Red Sea (?) and Amboina. If the synonymy were worked out I should probably include the Philippines, but no place further east as yet. Goniopora minor nom. n. for G. “ Great Barrier Reef (12) 5.'’ (Plate XL VIII, figs. 1, 3.) 1903. Bernard, p. 52, Great Barrier Reef (12) 5, pi. ii, fig. 6. Bernard’s specimen was “ only a chip " but corresponds well with No. 50, which is a smooth oval dome, with two little domes attached. The main mass is 8-5 cm. long and 4-5 cm. high, not in layers. The thecal tubes are continuous from top to bottom. The salient features are small calices, usually 2 mm. across, often 1-5 mm., deep for their size and open ; the 12 thick but narrow septa descending vertically, rudiments of third cycle sometimes visible, occasional junctions of septa before reaching the columella. Columella occupies most of the bottom of the calyx, and is much fused, but 4 to 6 septa run on to it and bear small spinulose pali. On the surface on the main dome these septa are thin and the pali very irregular, but on the small domes and at the edges of the large ones, they are always six in number, vertical and symmetrical, often joined by synapticula and making a low cylinder above the columella. These pali are borne by 6 thicker primary septa. The walls are thin, but appear thicker by being crossed at right angles by the septal edges. This is one distinction from Bernard’s G.B.R. (12) 8. Distribution : Known only from the Barrier Reef, at least until the synonymy can be worked out. Goniopora hirsuta sp. n. (Plate XLVIII, figs. 2, 4.) A minute colony which I should decline to describe were it not strikingly different from any other species in having the septal edges covered with fine compound spinules, giving them a hairy appearance. This was found on a mass of shells, Chama sp ?, heavily overgrown with lithothamnia,* dredged from Stn. XVI. It is a crust 3 mm. thick, yet in several layers, triangular in shape, 20 mm. along the side. It is probably one of those corals which never reach a much larger size, like Stylaraea punctata and Stglophora armata. Unfortunately only one-third of the area was alive when collected, the rest apparently overgrown by a crust of litho- thamnia, which is, however, so thin as to let the structures be seen. All calices are shallow, living calices 2-5 mm. across, the older ones, in the dead part, 3 mm. Walls of the former generally very thin, loosely made, and confused by compound spinules similar to those of the septa ; where walls are thicker septa cross them at right angles, and, in the older calices, this is always clear. Septa 12, equal ; sometimes from one to three fusions before the centre is reached ; and there are occasional rudiments of a third cycle. They are rather thick, and with their * I follow Stephenson, J. Linn. Soc. Zool. ‘ 1939 ’ p. 500. “ In this paper the word ‘ lithothamnia ’ (not italicised and without a capital) is used to indicate the encrusting Corallinaceae belonging to Litho- thamnion and related genera.” “ Lithothamnioneae ” is more correct, but is cumbrous, and implies the presence of more than one genus. 234 GREAT BARRIER REEF EXPEDITION spiimlose processes, tend to fill up the theca. The columella is visible only as one or two spinulose points. There are no pali. In the older calices pali are distinct, generally as six conspicuous points, hut they may be double, and such pairs may be connected to form a plate. The columella is rudimentary or an irregular fusion, bearing one or two upright spinulose points. In these older calices the septa are conspicuously thickened in and near the walls. Genus Alveopora. The long controversy over the relationships of this genus is given by Bernard (1903), pp. 4-9, and at length in J. Linn. Soc. Zool. XXYI, 1898, and XXVII, 1899. In those days morphology was occupying our minds rather to the exclusion of ecology, etc., and the attempt to make the epitheca and other coral parts into strictly defined morphological entities tended sometimes to unnecessary obscurity. In simple fact Alveopora is scarcely distinguishable from Goniopora, from which it is directly derived by reduction of building power, in walls and septa. Compare, e.g., Bernard’s (1903) division “ b ” of Goniopora (p. 179), calices with high thin walls, and note at foot of page, “at the sides of all these colonies the calices revert to the primitive type with thicker walls ; but for this they would be indistinguishable from true Alveoporae,” and other species have spinous septa also. (See, e.g., Bernard, 1903, pi. ix, figs. 1 and 2.) Yet in the Catalogue (p. 8) he maintains the position he took up in his earlier papers, in which, while agreeing that Goniopora is much like an enlarged Pontes, these two genera making the family Poritinae. Alveopora has no relationship, and belongs to another family. This arises from his obsession with the epitheca, which, when looked at simply, is seen to be merely an ecological adaptation, certainly not of the overwhelming morpho- logical importance Bernard attributes to it. With this Gardiner and Waugh (1939, p. 242) under “ Goniopora ” in the John Murray Expedition corals, agree. Alveopora irregularis, sp. n. (Plate XLIX, fig. 2; Plate L, fig. 1.) A single specimen, No. B.M. 607, dredged at Stn. XIV. Growth form columnar, with expanded top, the lower, constricted part tightly bound by a firm mass of Halimeda combined with sponge. This and the base of the expansion also wrapped in a stout wrinkled epitheca. A vertical section, to one side of the centre, shows no epitheca within the corallum, and the horizontal layers shown on Plate XLIX do not indicate pauses in growth, or increase by layer-over-layer formation, such as Bernard postulates in similar coralla of Goniopora. The species is very like A. allingi Hoffmeister from Samoa (1925, p. 81, pi. 23, figs. 2a, b and 2c) in the sizes of the calices, their walls and columellae ; but the septa, as seen in his fig. 2c, are distinctly different. Hoffmeister’s species has 12 nearly equal, distinct though narrow septa, not spines, of considerable breadth at the bottom, where they all join the columella. In A. irregularis they are 20-24 in number (in the larger calices) but most of them are low, toothed ridges or rows of very short, close-set spines, but, here and there, from one to three of them may broaden out into long spines. Such broader septa often meet in groups of two or three before joining the columella. The irregular network forming the columella is like that of A. allingi, but even broader, practically forming a MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA — CROSSLAXD 235 bottom to the cup in the larger calices. In the smaller it is of more normal size, and irregular in shape. Very few of the septa reach the top of the theca, though all approach it. It is possible that this species is merely A. allingi, which has suffered a degeneration parallel to that I have described (1931) in Leptastrea of Tahaiti. Alveopora mortenseni sp. n. (Plate XLIX, figs. 1, 3-4.) A single specimen was dredged from Stn. XXIII, but by the kind permission of Dr. Mortensen I am enabled to enlarge the description of the species by comparison with six dredged by him off Mauritius (Stns. 44 and 45, 15 and 16 Oct., 1929, the former from 25 fms.) Some of these are exactly like that from the Great Barrier Reef; others show variation in the direction of what Bernard (p. 101) calls “ that mysterious coral ” Porites clavasia, Aud. et Sav. of the Red Sea. While at Ghardaqa, Dr. Mortensen trawled a large number of corals of exactly this peculiar growth form, and it seems likely that an examin- ation of the series would connect this species with that of Savigny. These were from 30 fms. near Ashrafi lighthouse, at the entrance to the Gulf of Suez, and were accompanied by a number of species of gorgonians. As Plate XLIX, figs. 3-4 show, the growth of “ epitheca ” characteristic of the genus here reaches its greatest development. The branches lie loose on the muddy bottom, liable to burial, so that only their tips are alive, and even the dead stems would probably decay very rapidly if improtected. Calices round, variable in size up to 3 mm. There is a slight tendency for some on the sides of branches to be drawn out in the direction of the branch, which is more marked in some of the Mauritian examples. As the figure shows, buds are very numerous, wherever there is any space between calices there a bud develops. Rarely can a place be found where the walls are double, but they do occur, for parts of the circumference of a calyx. Otherwise the walls are extremely thin and highly perforate. The septal spines are, in this Barrier Reef specimen, so thin and irregular that no cyclic arrangement can be made out from them, but the number of short processes on the lip of the wall is twelve, indicating two cycles, and there are indications of a third ; within the calyx the upper parts of the walls bear a few short spines, scattered or more or less in vertical rows, lower down they become long, bending and fusing so that the original rows are quite confused (Plate XLIX, fig. 3). Generally they are very thin, almost hair-like, but thicker in calices low down in the colony. Deep down they meet and cross in the centre, and may form a tangle vaguely resembling a columella ; only in a few cases does this become a definite structure. Dr. Mortensen’s specimens might possibly be considered a variety, but having only one specimen from the Barrier Reef it is impossible to say whether the small differences observed are constant. While some calices are not distinguishable from those of the Barrier Reef specimen, the septal spines are usually thicker and more generally fall into rows ; sometimes a primary cycle can be made out, in rare cases it is very distinct and alone forms a definite columella. A tertiary cycle exists only as small spines alongside those of the secondary cycle. As for columellae, one specimen and two scraps have none, another and a scrap none or rudiments ; in a third thick primaries and some secondaries quite often make a distinct columella. 236 GREAT BARRIER REEF EXPEDITION The single small specimen from Mauritius (Stn. XLIV, 25 fms.) is even more delicate than the others ; the fine long spines rise from the walls at a higher level, making the theca appear shallower, though most are extremely irregular, some are in definite rows. Distribution : Besides the Great Barrier Beef only from Mauritius, and possibly, the Bed Sea. Genus Stylaraea M. E. and H. For a discussion of the rightness of separating this genus from Porites see Bernard (1905), p. 1 1 . I resuscitate it against the opinions of both Milne Edwards and Haime and Bernard ; though agreeing that it is a modified Porites its peculiarities warrant the distinction. I do not agree that the specimens before me, any more than those of Klunzinger, show imperfection due to youth, and I believe that their small size is normal for the species, as it is in Stylophora armata (?), of which I saw many specimens, all very small, in Tahaiti. It is always small, like Klunzinger’s specimens, in the Bed Sea also. Stylaraea punctata Klz. (Plate L, fig. 2.) 1879. Klunzinger, p. 44, pi. v, fig. 27. The species is generally attributed to Linnaeus and Esper, but to go behind Klunzinger is to land in the quagmire described by Bernard which I do not propose to enter (Bernard, 1905, p. 11, and under Porites Red Sea 9 and P. molluccas 1, pp. 161 and 243). To alter the attribution to Klunzinger is the only way to find footing without adding to the confusion by giving a new name. Eleven specimens were preserved, all labelled “ Mangroves, Passage 62, Low Island, Miniature Favia.” The largest, and one of the smallest, are on a fragment of a heavy shell, completely riddled by sponge, etc.; the other six are on a living Melina* shell ; and three on an Area shell, accompanied by a small colony of Porites stephensoni sp. n. (p. 241). They are accompanied by minute colonies of Leptastrea and Cyphastrea. The colonies of Stylaraea form cushion-shaped crusts, with a thin line of “ epitheca ” appearing round their bases. The largest is 15 mm. across and about 7 high ; most are not much smaller, but the smallest is only 7 mm. across and has 12 calices and some buds. As in Klunzinger’s example, calices are of uniform size, just under 1 mm. across. There are some details of importance to add to Klunzinger’s description, “ Die Septa sind sehr schmal, springen sehr wenig vor, sind leistenartig oder trabecular,” but, as seen in Klunzinger’s photograph when examined under a lens, in many calices one, more rarely two or three, reach the columella at a high level in the calyx. Barely all six of the first cycle may be seen to join the columella low down, but generally such regularity, if existing, is obscured. Two cycles of septa can usually be distinguished among the 12 little ridges or beams at the top of the wall. Costal ridges frequently cross the walls. The columella is not always a single style ; it is often double, or, more rarely, ends in three small points ; in the former case it may be a flattened plate. It is to be noted that in very young buds the stylar columella is already prominent. * Dr- Bruun informs me that the shell is correctly named, but that Pedalion (Solander) is more generally used. MADREPORARLl, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 237 The resemblance to Stylophora mentioned by Klunzinger is very superficial. All the above variations can be made out by careful examination of Klunzinger' s photograph, but, as it is very small, and in my copy fading. I give another. This shows details such as wall- structure ; and the minute spinules of the septa and tops of columellae, which may never have been visible in Klunzinger’ s. These spinules are always present, but conspicuous only in the largest specimen. Walls are more loose and irregular in smaller examples, but the septa and their costal ridges always rise above it. Whether they reach across it to the next calyx or not in the largest specimen they are always thickened over the wall, very much as in Klunzinger’ s photograph. Distribution : Only known hitherto from the Red Sea. Earlier records, if any, are without locality. Genus Pontes. Stephenson and others (1931) refer fourteen times to the genus, well illustrating its adaptation to all habitats, from pools and passages into the mangrove swamp, in the “ Mangrove Park,’’ reef flat, the moat, seaward slope and anchorage.* S. Manton’s traverses of the moat, shown at the tops of her pis. i to iii, illustrate an extraordinary dispro- portion between Porites and Montipora ramosa, the latter being vastly predominant. In only one part of the moat is Porites noted by Stephenson and others as predominant. There is also a “ Porites pond ” at Three Isles, and an abundance of small colonies on Yonge and Batt Reefs. The only place where colonies of a normally large size are recorded is the reef at Lizard Island, where much of the reef and its outer slope are made of colonies two yards across. This, it should be remembered, is a lagoon reef. Dr. Umbgrove (1939, p. 10) notes how this preponderance of Montipora (both foliosa and ramosa) in the moat of Leiden Island, in the Bay of Batavia, resembles conditions at Low Isles, but here, in deeper water, “ the largest corals are doubtless enormous Porites colonies (according to Yerwey, 1931, p. 206, Porites lobata )”, and there are other interesting differences. Umbgrove and 1 agree that “ we are only at the very beginning of having an idea of the true distribution of species, which abound on one reef but seem absent in other reefs.’’ I have given large scale examples under Pavona and Coeloria, but of the reasons for this and other phenomena of coral-growth we have no idea whatever. All this distribution of Porites is quite unlike the reef-areas I know in Equatorial East Africa, the Red Sea, or Tahaiti, and even the Marquesas, where colonies 6 feet x 6 feet and larger are striking objects everywhere.f S. Manton gives details. She also found the genus ubiquitous, but never saw or drew a colony of any size, her largest being 30 inches across, found in deeper water (see her graphs on pis. iii, vii and viii). Her criticism of Wood Jones on the formation of flat-topped cylinders is borne out by everyone who has ever seen a coral reef, and Wood Jones should not be quoted in future. She also quotes Marshall and Orr on removal of sediment, which, they say, is always done more effectively by large polyps. Yet here we have a genus in which the polyps arc of the smallest, yet which is peculiarly resistent to muddy water. The explanation is, as S. * It is here marked with an asterisk as characteristic of vertical or overhanging surfaces, surely a printer’s error ; cf. S. Manton. f A colony of ordinary size is shown on pi. i of the writer’s “ Forskaal’s Collection of Corals,” Univ. Mus. Zool., Kobenhavn, 1941. Mr. Otter’s photograph in P.Z.S., 1935, p. 504, shows an extensive growth of Synaraea convcxa in a Tahaitian lagoon, but the large colony is by no means one of the largest. 238 GREAT BARRIER REEF EXPEDITION Manton says, that Pontes removes mud easily, while unable to deal with coarser sediment. Another is that laboratory experiments do not last long enough to be comparable with nature ; at Ghardaqa the death of corals damaged by sediment was spread over several years. Whether this ubiquity of the genus applies to any of its species we have not enough evidence to decide. In this genus the usual difference between lighter calices on the upper, and stouter built on the lower sides of colonies is particularly striking. The phenomenon is usually explained by saying that the former are growing the more rapidly, but this could only be true if the resulting mass is to be columnar ; it cannot be true in the case of a regular dome. The fact is that the vast majority of museum specimens of Porites, as of other corals, were collected from reefs drying at the level of low water spring tides, where upward growth has ceased, or is about to cease. I call to mind innumerable colonies of Pocillopora, in which this difference reaches its extreme, the upper parts of the branches resembling empty honeycomb, wherever the branches have reached as high as they can, and are deformed by this fact. We find no thinning of the walls in the apical thecae of Acropora, where upward growth is at its greatest. Dredged specimens of massive Pontes are rare ; we have one in this collection, and in this the difference between upper and lower calices is much less than usual. I suggest that what we need for the proper determination of species in Pontes are series of examples of (1) young colonies, such as that of P. stephensoni of this collection, only more of them and from various habitats, (2) larger colonies, such as those described by Gravier and Vaughan, and (3) the full-sized colonies, however large. The last would, of course, be examined by means of small samples, say 10 cm. across, of which a large number would be needed, 6 to 9 from each mass, viz., one or two from the top surface, and six or eight from the sides, from near the top and low down, the latter collected in many cases by means of a diving hood. Drawings woidd be made showing the relations of the colonies to the reef and other surroundings, to currents, waves and light, and the samples numbered to correspond with the positions on these drawings. This would mean handling a mass of material difficult to bring home to European or American museums, and to store on arrival. There would therefore be a preliminary examination on the spot, and a careful selection made of the samples to be brought home. In some species it might be found that there is little variation in response to conditions ; in others results seem likely to be quite surprising, and likely to lead to general conclusions. Experiments would, of course, go on concurrently. This task would be a heavy one, quite outside the programme of such an expedition as the present one, which was occupied fully in more useful ways ; I suggest it for such a biological station as that of Ghardaqa in the Ked Sea, where the semi -fossil corals of the raised reef would of course be included. Solid Forms. Pontes stephensoni sp. n. (Plate L, figs. 3-4.) 1918. Porites haddoni Vaughan, p. 197, pi. 87, figs. 1-16. 1931. Porites haddoni Stephenson, p. 129, pi. 1, figs. 1, 3-4; pi. iv, fig. 1 ; pi. v, figs. 1-3, 5-9, 11. 1933. Porites haddoni Marshall and Stephenson, p. 236. The species is most readily distinguished from P. haddoni by the absence of the MADREPORARIA, HYDROCORALLIXAE, HELIOPORA, AXD TUBIPORA — CROSSLAND 239 columellar tubercle, which is never present unless as a mere rudiment, and by the proportions of " the triplet.” Six of the specimens were some of those used by Stephenson and Marshall hi experi- mental breeding. Having this very special importance I describe these specimens first, and then five others of the same species. These six are Nos. B.M. 232, 262, 305, 373, 377 and 381 . They are all very small, columnar in shape, but narrower at their bases and tops ; the largest 85 mm. high x 70 nun. in greater diameter ; the smallest 30 nun. high x 55 nun. in diameter. The narrower tops are bluntly pointed in four specimens ; No. 373, though narrower, is flat on the top. No. 262 has grown abnormally : the surface has small monticles, indicating that the fused columns of the colony are smaller than usual. There is much variation, No. 377 being covered with small, upwardly directed mounds, while No. 381, the other extreme of this series, has them much lower, broader, and therefore fewer. These growth forms are similar to those in Bernard’s (1905) pi. xii, fig. 4, or to those in Quelch’s (1886) pi. xi, fig. 8, which are very like Nos. 232 and 381 , but No. 377 is unique, and is there- fore illustrated on Plate L. The surface appears smooth to the naked eye but for a polygonal reticulation due to the thin walls being raised a little above the level of the contents of the calices, which appear shallow through being filled up by the septa and pali ; they are, in fact, when seen under the binocular, distinctly deep except, of course, those near the bases of the colonies. Taking No. 381 as an example, the walls are very thin, much perforate, and bear large granules, generally one above each septum, and flattened usually in the plane of the wall. The septa originate well below the edge of the wall, and bear a granule near it ; here may be an incomplete synapticular ring. The septa may run nearly horizontally to the pali, the free part of which is then comparatively short, or, more typically, the septa dip, leaving the pali as long, free, slender pillars. Sides and edges of septa, and especially the tops of the pali, finely spicular. Septa thin, much thinner than the loculi. The central fossa is comparatively wide and open, the septa being narrow, and there being no columellar tubercle, or, in some calices, only a microscopic rudiment. The top of the columella therefore forms a flat floor to the fossa, only sometimes the synapticular ring rises a little above its margin. It is generally solid, or nearly so. This clear, deep fossa without columellar tubercle is very characteristic. Gravier’s description of his P. somaliensis (p. 81) is very like this species : “ C’est cette taille relative- ment tres grande des palis qui constitue l’une des characteristiques les plus frappantes de ce Porites. Dans la plupart des calices la saillie columaire est indiscernable.” His figures, however, show a tubercle fairly frequently, and in Vaughan’s (1918) specimen it is generally present, though small. Gravier’s statement, taken alone, would apply equally well to P. lutea, as described by Klunzinger (1879, p. 40) and myself (1941, p. 24), but the pali of this latter are remarkably thick, and they hide the columella by being close together instead of leaving a wide fossa. The septa of “ the triplet ” in P. stephensoni are narrower than the others, the laterals often half their width ; they may be mere ridges, and joined, if at all, only deep down, immediately above the columella, where they meet the synapticular ring ; generally they are quite free. Pali on the paired septa are always tall, those on the single septum smaller ; on the “ triplet ” there may be one large one, three small ones, two small ones or none at all. In No. 377 the arrangement is the same as described above, but a small tubercle is sometimes present, and often the columella is looser, its radiating parts joining the synapticular ring like the spokes of a wheel. 240 GREAT BARRIER REEF EXPEDITION No. 373 is a colony with, a flat upper surface : here the pali are particularly well separated from the septa, and on some of the pairs they are “ V ” -shaped. A considerable area low down one side has shallow calices, thick septa, and inconspicuous pali. Some septa join over walls. Columella is solid, and there are no more traces of the columellar tubercle than in No. 381. No. 262, an abnormally shaped colony, may have grown on a vertical surface. Septa and walls often thicker than normal ; near the base much more so than in any other specimen. Columella solid. No. 232 is lightly built, so some wheel columellae are present. Laterals of “ triplet ” very narrow, but joined low down oftener than usual. No trace of a columellar tubercle. This specimen is labelled by Matthai (but not in his hand) “ Porites (? P. haddoni ).” No. 305 has deep calices ; thin, narrow, smooth septa ; very slender pali, spinulae at tops ; triplets mere ridges, and usually only four pali. Yet the walls are rather thick. Stephenson, p. 130, remarks : “ It is not impossible that more than one species provided us with planulae. The identification of Porites down to species is probably impossible in the field, and in any case there is no certainty as to what characteristics constitute specific limits in the genus, the matter being still within the realm of personal opinion.” I have described this series in detail since, so far as only six samples go, it shows that only one species was dealt with, and that a particularly well-defined one. It is of interest to compare the parents with the very young colonies bred by Stephen- son, and illustrated on his pi. v. There is hardly any resemblance, except in the case of fig. 12 (a “ colony of Pontes from a larva, or larvae, which settled on clean materials fixed in a box in Porites pond ”), which shows the conspicuous pali, single septal granule, and spinulation of this species. The chances are, therefore, that it belongs to this species, and that the differences from the other figures are due to age. It is interesting to note that the ease with which young colonies of Pocillopora and Porites , and of these genera only, are obtained artificially is paralleled by the fact that corals found on artificial objects, e.g., buoys and their chains, are nearly always species of Pocillojpora ( bulbosa or damicornis ), or more rarely Porites. In particular I may mention tiles and clean pearl-oysters put down for the purpose of obtaining young colonies in a bed of a great variety of corals at Dongonab ; in the course of a year or more only these two genera appeared. Pearl-shells with attached corals are conmon in museums and curio- shops, but I found in Dongonab that j of the hundreds of thousands of pearl-shells cultivated a proportion of which spent four or five years among growing coral, very few indeed had young coral colonies on them. It is a mystery, as it is difficult to believe that the substrata were unsuitable, for the other conditions were those in which many species of corals flourished. Besides the breeding series we have the following : No. B.M. 233 is a minute scrap from Batt Reef, Patch No. 1, labelled P. somaliensis by Matthai. It is a little coarser than those of the series, but has no better development of the columella, and the fossa is quite open. The “ triplet ” is, perhaps, better formed, and fusion more frequent. These conditions can be matched on many areas of the series. The absence of a columellar tubercle, and the usually free ends of the “ triplet,” divide it distinctly from P. somaliensis. In sample 122, a small scrap, the “triplet” is usually free, sometimes fused to a (synapticular) membrane. In places small columellar tubercles are frequent. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 241 Sample No. 4 is similar to No. 122 in calicular characters, but the “ triplet ” is more often fused to a part of the synapticular ring. The corallnm is a wholly detached nodule, covered everywhere with living calices, somewhat flattened, 45 mm. long, 37 mm. broad, about 27 mm. thick, with a somewhat undulating surface. It would be expected that movement of the corallum would bring every part into the same relationship to its surroundings, but the differences characteristic of the tops and bases of fixed colonies are found here, viz., at one end and on one side are found deeper, thinner-walled calices and long, slender pali and thin septa ; at the other end and other side distinctly shallower calices, stouter septa and shorter pah. No. 62, B.M. 368, along with small colonies of Stylaraea punctata is encrusting an Area shell from a mangrove passage. It has the usual thin perforated walls (they thicken below) besides the thick synapticular ring. This, and the large flat-topped columella- complex, close up the theca so that the loculi below its level appear as a ring of nearly round holes. The septa are fairly thick and coarsely granular on their sides, and the pali less conspicuous than usual ; these are features of the shallower calices. Over the umbo of the shell the crust is thicker, the calices are deeper, etc., but the great, flat columella is found over the whole colony, though it is not always quite solid. Sample 248 is one of the larger specimens, a nearly smooth column, attached by a narrow base, and narrowing again at the top ; 8 cm. high, about 6-5 cm. at the thickest. The calices, etc., are typical. One side of the colony has delicate calices with very narrow triplet-septa ; on the other side both are stouter. There are some few columellar tubercles near the base, but they are rudimentary. Eleven specimens with such constant characters justify my creation of a new species, though I do so with reluctance. Examination of a larger series, including older colonies from varied habitats, might prove the deep, open fossa to be a phenomenon of youth. It is conceivable that there was a columellar style or plate which has been reabsorbed during the breeding stage* to make room for the planula when the polyp is retracted, and the “ triplet ” may have been affected also, but it is all very unlikely. Porites haddoni Vaughan. 1918. Vaughan, p. 197, pi. 87, figs. 1, 1 a and 16 (with synonymy). 1925. Porites lutea Hoffmeister, p. 73, pi. 21, fig. 3. 1905. Porites Solomon Islands, 6. Bernard, p. 86, pi. viii, figs. 5-6 ; pi. xi, fig. 6.f 1938. Porites lutea haddoni Eguchi, p. 385. Two larger specimens and a small nodule, Nos. B.M. 399, 406 and 231, belong to this species. The latter was named by Matthai, but it is less typical than the other larger ones, which are from Low Isles, 231 being from Batt Reef. The growth-form is markedly different in the two larger specimens. No. 399 is a piece broken from the edge of the top of a large solid mass, two fused and flat-topped columns being included, having a total width of 1 1 cm. The surface of both is undulating. No. 406 is a low summit from the top of a larger mass, the surface consisting of a number of mammillae ; it is perfectly shown by Bernard’s pi. xi, fig. 6 of his “ Solomon Islands 6,” which differs little, if at all, from this species. * Bernard (1905), p. 21. | The reference to pi. xi, fig. 6, is not given in Bernard’s text, but in the explanation of the plates. 242 GREAT BARRIER REEF EXPEDITION The walls are particularly delicate, and, in places, tall, their grannies usually flattened in the same plane, projecting but little, and so delicately frosted as to be almost smooth. To the naked eye they form a delicate network over the surface of the corallum as in P. stepbensoni. The great difference between upper and lower calices of the sides of the corallum illustrated by Bernard, pi. viii, figs. 5 and 6, is even more striking than usual, and the shallow calices are not confined to the actual base ; this may be a specific character. The correspondence with Vaughan’s description is good, except that, like Hoffmeister, I should describe the columella as a “ compressed granulated plate, but it is often styliform.” Also “ as a rule the pali are distributed according to Bernard’s illustration 3 B ” (i.e., in the diagram on p. 19), “ although they may be arranged as in 3 C.” I cannot agree that P. haddoni is a synonym of P. lutea until the stabilization of this species, demanded in my account of Forskaal’s Collection, is carried out. Hoffmeister’s fig. 3 corresponds to part of the Red Sea specimen attributed to Forskaal, but, for the present, the majority of calices must show the big, thick pali, etc., of my own and Klun- zinger’s figures before the name P. lutea can be applied. Distribution : Great Barrier, including Murray Island ; Samoa, Palao and Solomon Islands. Probably also Fiji and Ellice Islands. Porites lobata Dana. 1907. Vaughan, p. 196. A small chip from the surface of a larger colony (Batt Reef, Patch 1, B.M. 236) is thus labelled by Matthai. It would seem to be nearest to Vaughan’s form parvicalyx, the portion he described in the middle paragraph of p. 201. The larger calices are 1 mm. across. This specimen is peculiar in that the tall, but imperfect synapticular ring above the edge of the columella is generally joined by the septa before they meet each other, so that the characteristic Poritid fusions cannot be made out, and, as the pali are very slightly developed, the bilateral pairs cannot be distinguished by their presence. In fact, in most calices, the septa resemble the spokes of a wheel joined to a thick hub. There is the further complication that, instead of a columellar style or plate, the synapticular ring is crossed irregularly by several delicate beams in any direction, nearly level with the tops of the small pali. Only rarely is a style or plate distinguished among them. It is particularly unfortunate that this scrap is all we have of this interesting corallum. Distribution : Hawaii, Fanning, Fiji and Samoa Islands. Now Great Barrier Reef. Porites solida (Forsk.). 1918. Vaughan, p. 191, pi. 84, figs. 3, 3 a. 1941. Crossland, p. 21, pis. i to iv. There are only four samples of this species, usually so abundant, viz., a small flat nodule, and a small piece of a crust tied together as sample 404, while Nos. 46 and 50 are pieces of the top edge of larger colonies. Distribution : Red Sea (the Arabian towns of both sides of the sea are built of it) ; Cocos Keeling, where Wood Jones says it is very abundant. It will probably be found abundant throughout the Indo-Pacific when the synonymy, and confusion with P. lutea, are cleared up. MADREPORARLA, HYDROCORA LLTNAE, HELIOPORA AND TUBPPORA— CROSSLAND 243 Porites australiensis Vaughan. 1918. Vaughan, p. 194, pi. 85, figs. 4, 4a, 5, 6 and 6a. (This gives also Bernard’s reference numbers.) A small nodule (Batt Beef, Patch 1, No. B.M. 234) is thus labelled by Matthai. It corresponds well with Vaughan's description and figures. Distribution : Great Barrier Reef and Murray Island ; Philippines. Pontes sp. indet. Sample 47 is a lamina, 7 cm. across the base, proximally 7 mm. thick. The calices recall those of P. haddoni, but I do not attempt identification of such a fragment, since I noted in Tahaiti (though not elsewhere) that both Synaraea species and solid species of Porites often form basally lamellar outgrowths, the calices of which vary greatly from those of the main masses. Porites lanuginosa Studer. (Plate LI, fig. 1.) 1901. Studer, p. 423, pi. xxix, fig. 9. 1907. Vaughan, p. 209 ; pi. lxxxvii ; pi. lxxxviii, figs. 1-la. The specimen dredged at Stn. XXV is rather larger than Studer’s, being 17 cm. x 11 cm. above, and 19 cm. high; 7 cm. was living, the rest being a narrower dead stalk, as in Studer’s specimen, but larger. The broken attachment is only 3 cm. wide and 11 cm. long ; the dead part is rotted by the usual organisms, and its original surface removed : it is partly covered by lithothamnia, Gypsina, and a thin crust of Montipora. Studer does not give the sizes of the calices, and from Vaughan’s figures they seem to be 1-6 mm. across ; but this refers to a specimen supplied by Verrill. I therefore hesitated to identify this Barrier Reef specimen, in which the calices are only 1 mm. across, and many, even on the humps, rather less. But Studer’s figure, being a colotype, is clearer than Vaughan’s half-tone, and agrees with this example quite clearly, showing too that many of the walls are not so thin as in Vaughan’s fig. 1 a. Studer gives no illustration of calices, which I therefore supply, attempting to show the characteristic high, spinulate wall and septal granules which give the coral its “ woolly,” or better, velvety surface. The triplets are usually fused, and the laterals bend to the median, which bears one palus ; but they are often free and bear three pali, which may sometimes fuse side by side to a plate. Pali tall, palial synapticular ring not visible. Distribution : Hawaii (Laysan) ; and now Great Barrier Reef. Pontes evermanni Vaughan. 1907. Vaughan, p. 194 ; pi. 80 ; pi. 81, fig. 2. Sample No. 1 is, apparently, the half of a solid, conical mass, the vertical section being 13 cm. wide and 10 cm. high, with undulating surface — the whole very like the figure of P. solida on Plate I of my description of Forskaal’s collection : the low projections are the tops of fused columns. It is quite different from the specimen figured by Vaughan on pi. 80 (but see the last paragraph of his description, in which he describes a series of growth forms parallel to Forskaal’s of P. solida). This specimen is remarkable for the lightness of its walls and septa over most of the surface, in spite of the weight and solidity of the colony as a whole. The synapticular 244 GREAT BARRIER REEF EXPEDITION ring inside the wall is conspicuous, but often incomplete. The differences from Vaughan’s description are : (1) that the pains on the dorsal directive, always small, is often suppressed altogether so that there are generally 5 pali, not 6. (2) I cannot make out the forking of the septa against the wall, nor is it visible in Vaughan’s figure. (3) In the very delicate calices of the upper parts the pali are often radially flattened. The two ends of the specimen differ markedly in the strength of the walls and septa — another case where some ecological details are called for. Distribution : Hawaii and now Great Barrier Beef. Porites densa Vaughan. 1918. Vaughan, p. 201, pi. 89, figs. 2, 2a and 26. Sample Nos. 414 and 408 correspond exactly with the above description, except that, in some calices, there are definite pali. No. 408 is a small, low mass 5 cm. in longer diameter and about 3 cm. in greatest height. No. 414 is broken from a solid mass, the broken surface 7 cm. thick and 9 cm. wide, not made of fused columns. Only an oval area, measuring 4 cm. x 7 or 8 cm., was living, the rest being covered with lithothamnia. In both cases the top is smooth, as in the type. Distribution : Great Barrier Reef and Murray Island only. Branched Species Sidnie Manton (1935) refers to them on p. 289 and in pis. vii and xi. Specimens are small and rare. Porites andrewsi Vaughan. 1918. Vaughan, p. 203, pi. 91, figs. 1, la, 2 and 2a. 1925. Porites andrewsi Hoffmeister, p. 77, pi. 22, figs. 2a-2c. Bernard’s “ Queensland 12 ”, “ Fiji 1,” “ Tonga 8 and 9,” and “ Solomon Isles 9.” Vaughan gives Bernard’s “Fiji 1 ” and “ Tonga 8 and 9 ” as “ almost certainly ” this species ; I omit the “ almost ” and add “ Solomon Islands 9,” which was overlooked by Vaughan, probably owing to an error in Bernard’s reference to the plates. In the text (p. 89) he refers to pi. xii, fig. 1, in the Explanation of the Plates to pi. xi, fig. 7. The former corresponds to the description of “ Caroline Islands 3 ” (p. 94), the latter to the description given of “ Solomon Islands 9,” and is typical of this species. There are three small fragments : samples 412-416 ; and a single branch dredged from Station XXVII that was tied to the lamina I have named Porites ( Synaraea ) vaughani, but not related to it : I take their being tied together to mean only that they are from the same station. Sample 412 is a small piece consisting of a fork, each branch with three small branch- lets ; it corresponds with both Bernard’s and Vaughan’s descriptions, but the growth form is more like Bernard’s figure. As Vaughan found, the peculiar arrangement of the “ triplet,” shown in Bernard’s diagram “ H ” on p. 19, is only occasional. Sample 416 is a broad fusion of two flattened branches, of equal size at the base, one of which aborts while the other grows on and branches. This branch bears low mounds MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 245 on one side (possibly the beginnings of branches) and two short branches on its other side, which demonstrate their readiness for fusion (characteristic of the species) by attaching themselves to a piece of dead, lithothanmia-covered coral. The dredged specimen is interesting in comparison with the one Mayer placed in “ deep water " (about 5 fins.) where there was no appreciable current (Hoffmeister, 1925 p. 77, pi. 22, fig. 2 c ; Mayor, Memorial Volume, 1924, p. 55, and table 8, p. 71, pi. 19, fig. 24d). This branch is similar, i.e., thin and round, but the calices are normal, almost cpiite level with the surface. A considerable part of one side is morbid, covered with broad flakes, among which calices are barely distinguishable. Clearly the habitat was unsuitable. Distribution : Great Barrier Beef ; Tonga, Fiji, Solomon and Samoa Islands. Porites suppressa sp. n. (Plate LI, figs. 2 and 3.) Xo.BAI.371 (" Xigger Head, Undine Beef. Branched massive Porites ”) is illustrated on Plate LI . The flat, much divided branch to the right is in marked contrast to the simpler, rounded, and more upright ones to the left ; and, in view of the importance attached to growth-form by some authors, a further selection of branches would have been interesting. I take it that the word " massive ” on the label, usually meaning “ solid,’' here means that the colony was large. The specific name refers to the fact that of the very short, thick septa several may be partly or wholly suppressed. The walls have thin edges, covering the surface with a reticulation, as in several other species ; the calices are filled up by pali and columella ; but, except near the base, they cannot be called superficial. The walls are thickened below by broad synapticular rings, which bear thick, spinulose septal granules close to the wall much larger than those of the wall, which are upright, thin, and often circumferentially flattened. Below these large granules the thick septa lie rather deep in the calyx, and are very short, joining a thick columcllar ring, so that it is often impossible to make out the usual fusions. That several septa are suppressed is obvious. With care one can determine that these may be one of a lateral pair, one or both of the laterals of the “ triplet,” or the dorsal directive, the suppression may be partial or complete. It appears as though some lateral pairs were fused throughout their lengths. The sides of the septa are rough. There are usually five tall pali, with rather swollen, roughened (not spinulate) ends, symmetrically surrounding a lower, but thick, prominent, and round columellar tubercle — thicker than in most species. A sixth small palus may be present on the dorsal directive, set back from the ring of the larger. Loculi appear rather as holes than slits, an outer circle between the synapticular rings, and about four minute holes between the inner ring and the columella. This describes the more open calices of the upper parts of the branches. Basally the walls and all granules thicken, the calyx becomes much shallower, and the loculi are obliterated, often so completely that the septa are invisible, except perhaps a low ridge or two. This thickening and shallowing extends over much larger areas on one side of the specimen than on the other. This species resembles Bernard’s “ Tonga 10 ” (p. 41, pi. ii, fig. 8 ; pi. xi, fig. 1) and, as Bernard himself remarks, “ Singapore 7 ” (p. 187, pi. xxviii, fig. 7 ; pi. xxix, fig. 1 ). Bernard separates these specieson account of the difference vt, 3. 17 246 GREAT BARRIER REEF EXPEDITION in growth forms, which seems to me quite insufficient. They differ from P. suppressa in having (1) smaller septal grannies near walls ; (2) often 3 pali on the triplet, as shown in the figure ; (3) small columellar tubercle ; (4) more distinct septa ; (5) “ the transverse section shows immense development of the horizontal elements.” Sections of the new species show nothing of this. Porites annae nom. nov. (Plate LII, figs. 1 to 3.) 1905. Pontes “ Great Barrier Reef 26.” Bernard, p. 129, pi. xvi, figs. 8 and 9 ; pi. xx, fig. 1. I am placing four specimens in this species, though only one corresponds closely with Bernard’s” Great Barrier Reef 26. ’’They are “ Batt Reef, Patch 1, Square 11 ’’(the specimen just referred to) and Samples 213, 4, and 15. Bernard’s illustration, pi. xx, fig. 1, is of a far better grown example than any of these, which are, in comparison, stunted. They show the same fundamental structure, though it results, in two cases, in striking variations of form. I therefore amplify Bernard’s description. The base of the colony is an incrustation, not free at the edges, or only rarely and to a trifling extent free, which sends up small, columnar branches usually with rather thicker ends. (Measurements may be taken from Plate LII.) These may fuse basally or higher up, in the latter case enclosing cavities. In the Batt Reef specimen and in No. 213 these branches are hardly more than knobs, whilst in No. 4 some attain a length of up to 35 mm., though the longest free branch is only 15 mm. In No. 15 some are regular columns 40 mm. high, but of these four, with some smaller, are completely fused by their sides to form a nearly straight-topped plate. In the Batt Reef specimen there is a fusion at one end into a solid mass, the columns composing which are visible on the surface only to the same extent as in normally solid growths. The difference between the calices of the tops and those of the sides of the branches is not always so marked as in Bernard’s description ; fairly deep and thin-walled calices may extend down the sides, though those on the basal crusts are always superficial. I do not find eight pali constantly : this number is only made up when there are three . on the “ triplet ” and one on the dorsal directive, besides the usual four on the lateral pairs ; and, in all the cases I have seen, the “ triplet ” bears three pali only occasionally (it may bear only one in the very next calyx) and the dorsal palus may be rudimentary or absent. The walls “ here and there surge up into a loose filamentous reticulum ; it is this which causes the colony to rise into branches.” In my view the appearance of this reticulum is the first stage in budding — which may be the same thing. “ The septa are thin and irregular in shape and position ” (this is not apparent in Bernard’s fig. 8) ; the pali, on the other hand, form a conspicuous ring of eight small, frosted points or plates.” In these specimens these exceptionally lightly made calices are not found on the tips of all the branches ; they are present, but rather exceptional, and are clearly a stage of growth ; in many cases the septa are regular, fairly thick, and with well-formed rounded pali even on the tops of branches. On their sides they are always like this, with pali conspicuous throughout, and, on the shallow calices of the basal crusts, remarkably thick and spinulose. Distribution : Great Barrier Reef of Australia only. MADREPORARIA, H YRR OCOR AXLEXAE , HELIOPORA AND TUBIPORA— CROSSLAND 247 Subgeniis Synaraea. Pontes ( Synaraea ) vaughani sp. n. (Plate LIII, figs. 4-5.) From Stn. XXYII (dredge) a free, squarisli lamina, 5 cm. across, about 5 mm. thick at one side, which is dead, very thin on the others, covered underneath with epitheca concentrically ridged, partly overgrown with sponge, polyzoa, etc. The main peculiarity of the species is that the broad coenenchymal ridges also bear calices, though more widely separated than those in the valleys. For details I refer to the plate. This shows that the calices in the valleys do not fall into rows, and are generally a little under 1 mm. across, while those on the mounds are a little wider. Those in the valleys are the more distinct because their loculi are more open and deep as well as nearer together. The coenenchyme forms rounded ridges, about 1 mm. high, finely spongy, bearing numerous, small, upright, spinulate processes. Septa very thick in all cases, spinulose on sides as well as edges, thus closing up the loculi still further. There are two granules on each septum between wall and palus, but they are often confused in the general spinulation. In spite of this the typical arrangement of the septa is easily seen ; the “ triplet ” is free, and bears one to three small pali, or none ; the triplet septa are shorter than the others. All pali low, but thick ; spinulose, generally distinct, but often no more than an emphasis of the general spinulation. The central fossa is open, and there is no columellar tubercle. I am taking this to be the expanded base of a branched corallum. It is close to Bernard’s “ Great Barrier 9,” “ Xorth AVest Australia 4 ” and “ China Sea 10 ” which, when more material is available, may turn out to be all the same, in spite of wide difference in growth form of one of them. The distance between Macclesfield Bank in the China Sea and Low Isles is now known to be not at all too great for them to be the homes of the same species, as was thought possible in Bernard’s time. I am naming this species after Dr. T. Wayland Vaughan, as a slight recognition of the immense and fruitful work lie has accomplished, especially in this genus, of which he is, in fact, the pioneer investigator. Porites ( Synaraea ) hawaiiensis Vaughan. 1905. Porites “ China Sea (19) 3,” Bernard, p. 167, pi. xxv, fig. 9. 1907. Porites hawaiiensis Vaughan, p. 216, pi. xci, figs. 2, 2 a. 1925. Porites ( Synaraea ) horizontalata Hoffmeister, p. 80, pi. 23, figs. 3«-36. Two very small, but complete, colonies were dredged from Station XXIV. They form thin crusts completely attached to substrata not now recognizable owing to decay. Their outlines are irregular, 40 mm. and 14 mm. across. The surface of the larger specimen is smoothly undulated ; whether this is due to irregularity of the substratum cannot be said. The scattered calices would be inconspicuous but that the thick walls and coenen- chyme rise a very little between them : it does not rise up into ridges. The calices are about 1 mm. in diameter, with about the same between their walls ; they are almost completely filled up by the thick septa, even the central fossa being only a minute hole — too small to allow any columella to be seen. The septa fuse in the usual ways, the lateral pairs and the central of the triplet ” bearing the five conspicuous pali. The triplet is fused, but occasionally one of the laterals may not reach the fusion of the 248 GREAT BARRIER REEF EXPEDITION other two. The pali are often not more conspicuous than the septal and mural “ granules,” which are upstanding pillars of much the same shape, and, like them, well spinulate. Details of wall structure cannot be made out ; the space between loculi of adjacent calices is a spongy mass, in which thick, horizontal elements bear upright processes, spinulate like those on the walls and septa. This seems to be a distinctive character, dividing the species from some described by Bernard. In some places these processes are smoothed away, not by abrasion, but by some morbid process by which the reticulum is thickened almost to solidity — far beyond what is shown in a normal section, where the horizontal elements are seen to be very thick and solid. The minute second specimen is much smoother, in fact as perfectly smooth as any Pontes. The calices are nearer together, and would be quite invisible to the naked eye if the walls did not rise very slightly. The pali are less conspicuous, tending to merge with the other spinulose granules. Otherwise it is like the preceding. These specimens are the encrusting, first growth-stage of the species described by Hoffmeister, which is the reason for giving an independent description. It is also possible that the Samoan specimens are the young of larger, possibly branching, species (c.f. the note on Pontes sp. above.) It is to be noted, however, that Hoffmeister had three specimens all, like these, dredged. They are also the young of the species described by Vaughan and Bernard, differing in that the ridges characteristic of Synaraea are present only as traces in the larger Barrier Reef specimen, and are absent in the smaller — a difference due only to youth. In his text Vaughan gives the diameter of the calices and their distance apart as only 0*5 mm., but both his figures show 1 mm. The columellar tubercle also is rarely seen in these specimens, more often in the smaller. It is in all cases minute. Distribution : Macclesfield Bank ; Hawaii ; Samoa ; and now Great Barrier Reef. Always dredged. H YDROCOR ALLIN AE . [Order Milleporina.] Genus Millepora. Stephenson and others describe the wide distribution of Millepora from the mangrove swamp to the reefs. S. Manton also finds it widely distributed on the reef flats to the crest of Yonge Reef, but of large size only below low water (p. 298). On the third 100 feet of the slope of Traverse III (to windward) “ Millepora forms very large branching colonies ” and, further out,” other corals identifiable from a boat are luxuriant, such as large Monti- poras, branching Millepora, etc.” Otherwise, judging from her scale drawings, the speci- mens seen are very small, except those on the coral head (pillar) of Plate XII, where a finely branched colony, 72 cm. across, and dome-shaped, occupies the summit. Apparently the big, upright plates of M. dichotoma or the solid honeycombs of M. platyphyllia do not exist on the Great Barrier Reef, nor any of the Tahaitian or Marquesan forms. Hickson (1898, p. 256) gave it as his opinion that all the species of this genus are really one, though, in conversation, he said that he could generally say where any given specimen had come from. Vaughan (1918, p. 206), after quoting Hickson says, “but it is at least convenient, if not systematically sound, to recognize by distinctive names the different aspects presented by colonies.” In 1941 I gave reasons for thinking that there are three species, genetically MADREPORARIA, HYDROCORALLINAE, HELIOPORA AND TUBIPORA— CROSSLAND 249 distinct, in the Red Sea, though I am still unable to decide whether the Tahaitian forms I described (1928, p. 727) are one or several species. M. alcicornis, of the West Indies, I believe to be distinct. So far as the small samples from the Barrier Reef are evidence I believe we have two or three species here, one of which is the first species in which a distinction has been based upon the actual structure of the corallum. Millepora tortuosa Dana. (Plate LIII, figs. 1 and 2.) 1848. Dana, p. 545, pi. 52, figs. 3, 3 a-b. The four small pieces present might possibly be taken for early stages of colonies of M. dic,h ot am a F., but, as far as my recollection goes, growths of this size would show the beginnings of regular reticulate plates, whereas these show no regularity at all. For the growth-form of a full-sized colony we have the diagrammatic figure given by S. Manton on Plate XII, which shows a low dome, 81 cm. across, covered with comparatively short branches (i.e., the form is truly cespitose). According to Dana the small size of the pores is diagnostic. I find here that in some specimens the pores are quite large, but all agree in being remarkably shallow, and therefore inconspicuous. The variation in their numbers is large, but this is often the case in other species. Specimen No. 189 shows an unusual regularity of form, but is in no way different from the others. No. 409 is a solid plate with few irregular branches, much infested by cirripedes. No. B.M. 427 may or may not be a distinct species. It is broken to pieces, but the illustration shows its peculiarly loose and round branches. It was dredged from Stn. XXVII, so the habitat may account for these peculiarities. It agrees with the reef- specimens in the scarcity and shallowness of the pores. The species differs physiologically from M. dichotoma F. in becoming white on drying, as does the species, described by Vaughan from Cocos Keeling, to which he gives this latter name ; but M. dichotoma of the Red Sea keeps its rusty brown colour after drying, and even after maceration. Distribution : It is impossible to give the distribution of any form of this genus. It can be said, however, that this one does not occur in the Red Sea, nor in Tahaiti, and, apparently, not cast of Fiji. Millepora foveolata sp. n. (Plate LIII, fig. 3.) Four specimens, P. 20, 31, 41 1 and 252, encrusting dead coral ; with a note that it often covers unattached pieces. They are now [12 . iii . 1942] of a light greenish colour, and all have a nodular surface, which, unlike “ M. verrucosa ,” is not due to parasites. The most nodular, P. 20, is infested with cirripedes, but they choose the flat sites between nodules, or, if they occupy a projection they actually flatten it and the area round them. No. 31 is much smoother ; over most of its surface it is undulating rather than nodular. Certain parts of it are infested with numbers of small polycliaeta, probably Spionids, as well as a few cirripedes ; but these have no effect upon the structure of the surface now to be described. The projections are not wart-shaped, but rather irregularly columnar, as shown on Plate LIII. Similar encrusting growths of the same colour and smooth or not are common every- 260 GREAT BARRIER REEF EXPEDITION where, abounding in the northern Red Sea for instance ; but the present ones differ even to the naked eye, to which they seem to be crowded with large pores. Under a lens this is seen to be due to each pore, or a short row of dactylopores, being at the bottom of a funnel- shaped depression formed by a sharp ridge. Dactylopores surround gastropores in the usual way ; the former are more numerous than usual, and often two circles round one gastropore can be made out. The ridges do not surround a set of gastro- and dactylopores, but each separately. No. P.20 is practically the same on all sides, but with smooth areas between the projections on one side ; No. 31 is completely foveolate on one side, most of the underside smooth ; No. 252, an irregular growth which does not extend onto the underside of its foundation, has considerable smooth areas between the projections ; No. 411, a very small piece broken from a fixed mass, is typical. Millepora sp. Possibly M. platyphyllia Dana, which is not Ehrenberg’s species of the same name from the Red Sea, as Hoffmeister has already remarked ; specimens resembling these of the Great Barrier Reef have not been seen in the Red Sea. They consist of three small pieces broken from the tops of small plates, ending in irregular, flattened, finger-shaped lobes, two or three of which may fuse to make a broader expansion. An interesting point is their heavy infection with a cirripede, the same as that found rarely in M. foveolata, which, unlike the case of “ verrucosa ” examined by Hickson and verified by me in the Red Sea, do not produce wart-like projections of the surface. Many make no alteration at all ; others are on low flat raised areas ; but it is doubtful how far these are caused or modified by the cirripede, as there is generally a mere film of Millepora over the cirripede, which has a quite flat, not conical, upper surface. It is impossible to give a name to these few pieces of such a general type of Millepora without an intimate personal knowledge of the Great Barrier Reef fauna. Hoffmeister may or may not be right in placing Dana’s platyphyllia and squarrosa var. incrassata under M. truncata Dana. This species is perfectly smooth, with numerous and conspicuous pores. It is white when dry. [Sub-class] Alcyonaria.* [Order Coenothecalia.] Heliojjora coerulea. The single specimen brought home, No. 58, is of the form in which solid plates end in slender vertical branches. Some of them are fused together, others free, as illustrated by Milne Edwards and Haime (1857a), pi. F. 1, fig. 3a. Another form, in which large radiating plates end in rounded lobes, is shown in Yonge’s Year on the Great Barrier Reef, pi. xvii a ; and in Gardiner’s (1931) Coral Reefs and Atolls, pi. xii. Hickson (1924, p. 118) says : The form of the colony is very variable. It may be branched like a stag’s horn Madrepore, laminate, or almost massive, but the ends of the branches are usually blunt and lobed. It sometimes reaches a size of three or four feet in diameter by two or more in height.” It * Crossland wrote “ Alcyonariidae” MADR EPORARLi, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND 251 is ci great reef builder, though usually best developed in lagoons ; witness the frequent references to it in Gardiner's descriptions of Indian Ocean and Pacific reefs. Distribution : Indian Ocean, from the Laccadives to the Seychelles, but not in the Red Sea. Present hi Malaysia and the Pacific to Funafuti ; not recorded from Samoa, and not present in Tahaiti. Tubipora musica. Stephenson and others describe this species as present almost everywhere, from Low Isles to the crest of Yonge Reef, on which S. Manton shows (pi. xiv) two large colonies, the larger 03 x 36 cm. As she did not meet it on the traverses it is presumably rare at Low Isles ;t it is generally more abundant and larger on outer reefs, always so in the Red Sea. For the specific identity of all varieties see Ilicksont (1924 p. 115). He examined many hundreds of specimens from the beach of Celebes, and many scores alive, and found that every variety known could be found on that one shore. It was sold in Arab drug shops in Jerusalem at least as late as 1918. Distribution : Northern Red Sea ; throughout the Indian Ocean and Pacific ; but not recorded from Samoa ; and not present in Tahaiti. REFERENCES. Alcock, A. 1902. Report on the Deep-sea Madreporaria of the Siboga Expedition. Res. expl. zool. bot. . . . Siboga, Monogr. XVIa, 52 pp., pis. i-v. Audouin, V. 1826. In : Savigny, J. C., Description de l’Egypt. Yol. 1, part 4. Explication som- maire des Planches de Polypes . . . , pp. 225-244, 14 pis. Baker, J. R. 1925. 40. A Coral Reef in the New Hebrides. Proc. zool. Soc. London, pp. 1007-1019, 8 text-figs. Bedot, M. 1907. Madreporaires d’Amboine. Rev. suisse Zool. Geneve, 15, pp. 143-292, 50 pis. Bernard. H. B. 1896. The Genus Turbinaria, The Genus Astraeopora. Cat. Madrep. Corals British Museum (Nat. Hist.) II, 106 pp., 33 pis. 1897. The Genus Montipora, The Genus Anacropora. Cat. Madrep. Corals British Museum (Nat. Hist.) Ill, 192 pp., 34 pis. 1903. The Family Poritidae. I : The Genus Goniopora. Cat. Madrep. Corals British Museum (Nat. Hist.) IV, 206 pp., 14 pis. 1905. The Family Poritidae. II : The Genus PoritfS. Part 1 : Porites of the Indo-Pacific Region. Cat. Madrep. Corals British Museum (Nat. Hist.) Y, vi -)- 303 pp., 35 pis. Bertram. G. C. L. 1936. Some Aspects of the Breakdown of Coral at Ghardaqa, Red Sea. Proc. zool. Soc. London, 1936, pp. 1011-1026, 1 pi. Boschma, H. 1923. The Madreporaria of the Siboga Expedition. Part IV : Fungia patella. Res. expl. zool. bot. . . . Siboga, Monogr. XVId, 20 pp., 2 pis. 1929. The Fungidae (Anthozoa) Collected by Mr. Cyril Crossland at Tahiti and Neighbouring Islands. Proc. zool. Soc. London, 1929, pp. 43-47, 1 pi. Brook, G. 1893. The genus Madrepora. Cat. Madrep. corals Brit. Mus. (Nat. Hist.) 1, 212 pp., 35 pis. Bruggemaxx, F. 1877. Notes on Stony Corals in the Collection of the British Museum. Ill: A Revision of the Recent Solitary Mussaceae. Ann. Mag. Nat. Hist. (4) XX, pp. 300-313. Crossland, C. 1928a. The Island of Tahiti [and discussion]. Geogr. Journal London, LXXI, pp. 561-585, text-illustr. 19286. Notes on the Ecology of the Reef-builders of Tahiti. Proc. zool. Soc. London, 1928, pp. 717-735. 1 text-fig., 5 pis. [For corrections see ibid., 1931, pp. 391-392.] t Y^onge (1930, p. 49) says “ absent.” Stephenson and others (1931, p. 71) say “ rare.” J Crossland, in his typescript, referred also to Hickson’s paper in Quart. J. micr. Sci. London, XXIII, 1883. But in it there appears to be no reference to the specific identity of all varieties of Tubipora. — [A.K.T.] 252 GREAT BARRIER REEF EXPEDITION 1931. Tlie Reduced Building Power and Other Variation in the Astrean Corals of Tahiti, with a Note on Herpetolitha Umax and Funqia spp. Proc. zool. Soc. London. 1931, pp. 351-392, 13 text-figs., 22 pis. 1935. Coral Faunas of the Red Sea and Tahiti. Proc. zool. Soc. London, 1935, pp. 499-504, 3 pis. 1939. The Coral Reefs at Ghardaqa, Red Sea. Proc. zool. Soc. London, series a, 108, pp. 513- 523, 3 text-figs. 1939ff. Further Notes on the Tahitian Barrier Reef and Lagoons. Journ. Linn. Soc. London, 40, pp. 459-474, 2 pis. 1941. On Forskaal’s Collection of Corals in the Zoological Museum of Copenhagen . . . Skr. Univ. Zool. Mus. Kjobenhavn, 1, 63 pis., 12 pis. 1948. Reef corals of the South African Coast. Ann. Natal Mus. XI, pt. 2, pp. 169-205, 10 pis. Dana, J. D. 1848. Zoophytes. U.S. Expl. Exp. 1838-42. . . . VII, 740 pp., 61 pis. — - — 1849. Atlas. Zoophytes. U.S. Expl. Exp. 1838-42. ... 12 pp., 61 pis. Doderlein, L. 1902. Die Korallengattung Fungia. Abh. senckenb. naturf. Ges. 27, pp. 1-162, 25 pis. Duncan, M. 1883. Observations on the Madreporarian Family — the Fungidae, with Special References to the Hard Structures. J. Linn. Soc. London, 17, pp. 137-162, pis. 5-6. Edwards, H. Milne. 1849. In : Cuvier, G. L. C. F. D., 1836-49. Le Regne animal . . . Paris, Third (“ Disciples ”) Ed. Tom. 20, Zoophytes, Atlas, 100 pis. col. 1857a. Hist. Nat. des Coralliaires, Atlas, pp. 11, 31 pis. 1860. Hist. Nat. des Coralliaires. Ill : Suite de la section des Madreporaires Apores, 560 pp. and Haime, J. 1848. Recherches sur les polypiers. III. Monographic des Eupsammides. Ann. Sci. nat. Paris (3) Zool. X, pp. 65-114, pis. 1. 1848. Recherches sur les polypiers. IV. Monographie des Astreides. Ann. Sci. nat. Paris (3), Zool. X, pp. 209-320, pis. 5-9. 1850. A Monograph of the British Fossil Corals. I. Introduction ; Corals from the Tertiary and Cretaceous Formations. Palaeontographical Soc. London, Monogr. — — 1857. Hist. Nat. des Coralliaires. II : Madreporaires, 633 pp. Eguchi, M. 1938. A Systematic Study of the Reef-building Corals of the Palao Islands. Studies Palao Trop. Biol. Sta. Tokyo, 3, pp. 1-390. Ellis, J., and Solander, D. 1786. Natural History of . . . Zoophytes . . . London, pp. xii + 208, 63 pis. Esper, E. J. C. 1797. Fortsetzungen der Pflanzenthiere, etc. Theil, I, 230 pp. Illust. Faustino, L. A. 1927. Recent Madreporaria of the Philippine Islands. Monogr. Bur. Sci. Manila P.I. 22, 310 pp., 100 pis. Fowler, G. H. 1885. The anatomy of the Madreporaria : I. Quart. J. Micr. Sci. London, n.s. C, pp. 577-597, pis. xl-xlii. 1887. The Anatomy of the Madreporaria : III. Quart. J. Micro. Sci. London, n.s. CIX (XXVIII), pp. 1-19, pis. 1-2. Gardiner, J. S. 1898. On the Turbinolid and Oculinoid Corals Collected by the Author in the South Pacific. Proc. zool. Soc. London, 1898, pp. 994-1000, pi. lxii. 1899. On the Solitary Corals, Collected by Dr. A. Willey. Willey, A., Zoological Results, Part II, pp. 161-180, pis. xix and xx. 1899a. On the Astraeid Corals Collected by the Author in the South Pacific. Proc. Zool. Soc. London, 1899, pp. 734^764, pis. xlvi-xlix. 1904. South African Corals of the Genus Flabellum , with an Account of their Anatomy and Development. Mar. Invest. S. A. II. pp. 115-154, 4 pis. 1904a. The Turbinolid Corals of South Africa, with Notes on their Anatomy and Variation. Mar. Invest. S. A. Ill, pp. 95-129, 3 pis. 19046. Madreporaria. Parts I and II. The Fauna and Geography of the Maidive and Laccadive Archipelagoes, II, Part III, pp. 755-790. Pis. lix-lxiv. 1905. Madreporaria. Parts III and IV. The Fauna and Geography of the Maidive and Laccadive Archipelagoes, II, Suppl. I, pp. 933-957, pis. lxxxix-xciii. 1909. The Madreporarian Corals : I. The Family Fungiidae, with a Revision of its Genera and Species and an Account of their Geographical Distribution. Rep. Percy Sladen Trust Exped. Indian Ocean, 1905. Trans. Linn. Soc. (Zool.) London, (2) XII, pp. 257-290, pis. 33-39. 1938. The Flabelloid and Turbinolid Corals. Sci. Rep. John Murray Exped. 1933-34, V, pp. 167-202, 6 text-figs., 7 pis. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AXD TUBIPORA— CROSSLAND 253 Gardiner. J. S., and Waugh, P. 1939. Madreporaria excluding Flabellidae and Turbinolidae. Sci. Rep. John Murray Exped. 1933-34. VI, pp. 225-242, 2 pis., 3 text-figs. Gravier, Ch. 1911. Les recifs de coraux et les Madreporaires de la baie de Tadjourah. Ann. Inst. Oceanogr. Paris, II, fasc. 3, 96 pp., 12 pis. Harrison, R. M., and Poole. M. 1909. Marine Fauna from the Mergui Archipelago, Lower Burma, collected by Jas. J. Simpson . . . and R. X. Rudmose-Brown . . . Madreporaria. Proc. zool. >Soc. London, 1909, pp. 897-912, pis. Ixxxv-lxxxvi. Hickson, J. S. 1898. On the Species of the Genus Miflepora : Proc. zool. Soc. London, 1898. pp. 246-257. 1924. An Introduction to the Study of Recent Corals. Manchester. XIV -j- 257 pp. 110 text- illustrations. Hoffmeister. J. E. 1925. Some Corals from American Samoa and the Fiji Islands. Paper Dept. Mar. Biol. Carnegie Instn. Washington. XXII, iv -f- 90 pp., 23 pis. 1929. Some Reef-corals from Tahiti. J. Wash. Acad. Sci. 19, Xo. 16, pp. 357-365, 2 pis. Horst, J. D. van der. 1921. The Madreporaria of the Siboga Expedition. Part II : Madreporaria. Fungidae. Res. expl. zool. bot. . . . Siboga, XYI6, 46 pp., 6 pis. 1922. Xo. IX: Madreporaria: Agariciidae. Trans. Linn. Soc. (Zool.) London (2) XVIII, Pt. 1, pp. 417—429, 2 pis. 1922a. The Madreporaria of the Siboga Expedition. Part III : Eupsammidae. Res. expl. zool. bot. . . . Siboga, Monogr. XYIe, 75 pp., pis. vii, viii. Kent, W. Saville. 1871. On some Xew and Little-known Species of Madrepores, or Stony Corals, in the British Museum Collection. Proc. zool. Soc. London, 1871, pp. 275-286, pis. xxiii-xxv. 1893. The Great Barrier Reef of Australia ; its Products and Potentialities, 387 pp., 64 pis., text- illustr. Koch, G. von. 1886. t'ber das Yerhaltnis von Skelet and Wcichtheilen bci den Madreporen. Morph. Jahrb. XII, pp. 154-160, tafel ix. Klunzingbr, C. B. 1879. Die Korallthiere des Rothen Meeres. II: Die Steinkorallen. 1: Die Madre- poraceen und Oculinaceen, 88 pp., 10 pis. 1879a. Die Korallthiere des Rothen Meeres. Ill : Die Steinkorallen. II : Die Astraeaceen und Fungaceen. 100 pp., 10 pis. Lamarck, J. B. P. A. de M. de. 1816. Histoire naturellc des Animaux sans vertebres, presentant les caracteres generaux et particuliers de ces Animaux. . . . precedee d’une introduction offrant la determination des caracteres essentiels de Tanimal . . . enfin Imposition des principes fondamentaux de la Zoologie. Tome II, 568 pp. Manton, M. S. 1935. Ecological Surveys of Coral Reefs. Sci. Rep. Gt. Barrier Reef Exped. Ill, Xo. 10, pp. 273-312, 2 text-figs., 16 pis. Marenzeller, E. von. 1901. Ostafrikanische Steinkorallen. Jahrb. Hamburg wiss. Anst. XVIII for 1900, pp. 117-134, 1 pi. 1904. Steinkorallen. Wiss. Ergebn. Valdivia Exped. VII, Lief. 3, pp. 261-318, pis. xiv-xvii. — — 1906. Riffkorallen. Denkschr. Akad. Wiss. Wien, 80, XXVI, Fortsetzung, pp. 27-97, 29 pis. 1906a. Uber den Septennachwachs der Eupsamminen. E. H. Denkschr. Akad. Wiss. Wien, 80, XXIV, Fortsetzung, pp. 1-12, 7 text-figs. Marshall, S. M., and Stephenson, T. A. 1933. The Breeding of Reef animals. Part I : The Corals. Sci. Rep. Gt. Barrier Reef Exped. Ill, Xo. 8, pp. 219-245, 1 pi. Matthai, G. 1924. Report on the Madreporarian Corals in the Collection of the Indian Museum. Part I : Mem. Ind. Mus. Calcutta. VIII, pp. 1-59, pis. i-xi. 1928. A Monograph of the Recent Meandroid Astraeidae. Brit. Mus. (Nat. Hist.) Cat. Madrep. Corals, VII, 288 pp., 72 pis. Mayer, (afterwards Mayor), A. G. 1918. Ecology of the Murray Island Coral Reef. Papers Dept. Mar. Biol. Carnegie Instn. Washington. IX, pp. 1-48, pis. 1-19. Mayor, A. G. 1924. Growth-rate of Samoan Corals. Papers Dept. Mar. Biol. Carnegie Instn. Washington, XIX. 4, pp. 51-72,26 pis. Moseley, H. X. 1881. Report on Certain Hydroid, Alcyonarian and Madreporarian Corals Procured . . . 1873-1876. Report Sci. Res. H.M.S. " Challenger ”... Zoology : II. Part III, On the Deep-Sea Madreporaria, pp. 127-248, 16 pis. Paradice, W. E. J. 1925. The Pinnacle- or Mushroom-shaped Coral Growths in Connection with the Reefs of the Outer Barrier. Trans. Roy. Geogr. Soc. Australasia (Queensland), I, pp. 52-60, pis. vi, vii, 4 diagrams, 9 sections. Quelch, J. J. 1886. Report on the Reef-corals Collected by H.M.S. Challenger during the years 1873- 76. Report Sci. Res. H.M.S. “ Challenger ”... Zoology, XVI, 203 pp., 12 pis. 254 GREAT BARRIER REEF EXPEDITION Savigny, J. C. [1809-1829] 1817. Description de l’Egypt Hist. Nat. Planches, Yol. 2. [Vol. 1, part 4, Explication des Planches de Polypes, pp. 225-244, 14 pis., 1826, by Audouin, Y. (no MS. by Savigny).] Stephenson, T. A. 1931. Development and the Formation of Colonies in Pocillopora and Porites : Part I. Sci. Rep. Gt. Barrier Reef Exped. Ill, No. 3, pp. 113-134, 6 pis. Stephenson, T. A., Stephenson, A., Tandy, G., and Spender, M. 1931. The Structure and Ecology of Low Isles and Other Reefs. Sci. Rep. Gt. Barrier Reef Exped. Ill, No. 2, pp. 17-112, 15 text- figs., 27 pis. Studer, Th. 1901. Madreporarier von Samoa, den Sandwich-Inseln und Laysan. Zool. Jahrb. Jena (Syst.) 14, pp. 388-428, pis. 23-31. Thiel, M. E. 1932. Madreporaria zugleich ein Versuch einer Vergleichenden Oekologie der gefundenen Formen. Mem. Mus. Hist. nat. Belg. Hors. Ser., 2, 12, 177 pp., 21 pis. Umbgrove, J. H. F. 1939. Madreporaria from the Bay of Batavia. Zool. Meded. Leiden, XXII, pp. 1-64, 4 text-figs., 18 pis. Vaughan, T. W. 1907. Recent Madreporaria of the Hawaiian Islands and Laysan. Bull. U.S. nat. Mus. 59, 427 pp., 96 pis. 1917. Some Corals from Kermadec Islands. Trans. Proc. N.Z. Inst. XLIX, pp. 275-279, pis. xvii-xx. 1918. Some Shoal-water Corals from Murray Island (Australia), Cocos-Keeling Islands, and Fan- ing Island. Papers Dept. Mar. Biol. Carnegie Instn. Washington, IX, pp. 49-219, 2 text-figs., pis. 20-93. and Wells, J. W. 1943. Revision of the Sub-orders, Families, and Genera of the Scleractinia. Geol. Soc. Amer. Special Pap. No. 44, pp. xv, 363, 51 pis. Verrill, A. E. 1863. List of the Polyps and Corals sent by the Museum of Comparative Zoology to Other Institutions in Exchange, with Annotations. Bull. Mus. Comp. Zool. Harvard, 1, pp. 29-60. 1868. Synopsis of the Polyps and Corals of the North Pacific Exploring Expedition, . . . 1853- 1856 . . . Additional Species from the West Coast of North America. Communications of the Essex Inst. Salem, V, for 1866-7, pp. 17-50, 2 pis. 1901. Variations and Nomenclature of Bermudian, West Indian and Brazilian Reef Corals, with Notes on Various Indo-Pacific Corals : IV. Conn. Acad. Arts Sci. XI, pt. 1, pp. 63-168, pis. x-xxxv. Verwey, J. 1931. Coral Reef Studies, II. The Depth of Coral Reefs in Relation to their Oxygen Con- sumption and the Penetration of Light in the Water. Ill : Geomorphological Notes on the Coral Reefs of Batavia Bay. Treubia, XIII, livraison 2, pp. 169-215, 1 pi. Waugh, P. 1936. The Variation and Species of Red Sea Turbinaria and Astraeopora, with a Discussion of the Genera. Proc. zool. Soc. London, 1936, pp. 913-929, 2 pis. Wells, J. 1936. The Madreporarian Genus Polyaster Ehrenberg. Ann. Mag. Nat. Hist. (10), XVIII, pp. 549-552, 2 pis. Wood-Jones, F. 1907. On the Growth-forms and Supposed Species in Corals. Proc. zool. Soc. London, 1907. Pp. 518-556, text-figs. 145-161, pis. xxvii-xxix. 1910. Coral and Atolls, London, xxiii + 392 pp., 79, text-figs, portrait, map, 27 pis. Yabe, H. and Eguchi, H. 1935. Revision of Reef Coral Genera Echinopora, Oxyphyllia, Mycedium , Oxypora, and Physophyllia. Proc. Imp. Acad. Tokyo, XI, pp. 429-431. Yabe, H., Sugiyama, T., and Eguchi, M. 1936. Recent Reef-building Corals from Japan and the South Sea Islands under the Japanese Mandate : I. Sci. Rep. Tohoku Univ. 2nd ser. (Geol.), special Vol. I, 66 pp., 59 pis. Yonge, C. M. 1930. Studies on the Physiology of Corals. I : Feeding Mechanisms and Food. Sci. Rep. Gt. Barrier Reef Exped. I, No. 2, pp. 13-57, 34 text-figs., 2 pis. MADREPORARIA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAND INDEX abdita, Favites . PAGE . 129 danae, Pocillopora abrotanoides, Acropora . 204 danai, Pavona .... Acanthastrea . 140 decussata, Pavona Acropora .... . 198 Dendrophyllia .... actiniformis, Fungia . . 152 densa, Porites .... aequicostatus, Heterocyathus . . 102 digitata, Montipora affinis, Acropora . 205 digitifera, Acropora Alveopora .... . 234 doreyensis, Favia ampliata, Merulina . 151 duerdeni, Pavona andreicsi, Porites . 244 angularis, sp. n., Montipora . . 197 echinata, Acanthaslrea annae, nom. nov., Porites . 246 echinata, Fungia arbuscula, Dendrophyllia . 170 Echinopora .... armata, Acropora . 212 elirenbergana, Leptastrea aspera, Acropora . 205 elseyi, Acropora .... aspera, Favites . . 132 erythraea, Montipora . aspera, Oxypora . 159 evermanni, Porites astraeiformis, Coeloria . 147 exesa, Psammocora Astraeopora . 180 exilis, Acropora .... australiensis, Porites . . 243 eydouxi, Pocillopora . benhami, Goniastrea . . 136 fascicularis, Galaxea . bifrons, Turbinaria . 176 favus, Favia .... bottae, Leptastrea . 116 favus var. crassidens, var. n., Favia broohi, sp. n., Acropora . 228 fimbriata, Euphyllia . bruggemanni, Acropora . 221 foliosa, Montipora cactus, Pavona . . 162 formosissima, Stephanophyllia fossae, sp. n., Montipora canalis, Acropora . 219 foveolata, sp. n., Millepora . cancellata, Acropora . 225 foveolata, Montipora . Caulastrea .... . 139 frondens, Turbinaria . chalcidicum, Cyphastrea . 117 fruticosa, Acropora clavigera, Acropora . 226 fruticosa, Montipora . clavus, Galaxea . . 122 fungites, Fungia Coeloria .... . 147 furcata, Caulastrea coerulea, Heliopora . 250 composita, sp. n., Montipora . . 195 Galaxea ..... contigua, Psammocora . 165 geojfroyi, Trachy phyllia corymbosa, Acropora . . 211 glabrescens, Euphyllia corymbosa, Lobopliyllia . 142 glochiclados , Acropora crispa, Oulophyllia . 147 Goniastrea ..... Crustacea, Podobacia . . 156 Goniopora ..... curta, Orbicella . . 124 grandis, Acropora cyclolites, Fungia . 153 granulosa, Montipora . Cynarina .... . 137 Cyphastrea . 117 haddoni, Porites daedalea, Coeloria . 148 haimei, Acropora halicora, Favites damicornis, Pocillopora • . 110 hawaiiensis, Porites ( Synaraea ) PAGE . 110 . 161 . 161 . 169 . 214 . 183 . 207 . 127 . 163 . 141 . 152 . 118 . 115 . 223 . 193 . 243 . 165 . 223 . 112 . 172 . 125 . 126 . 105 . 194 . 154 . 186 . 249 . 188 . 176 . 220 . 184 . 153 . 140 . 121 . 101 . 104 . 213 . 132 . 231 . 202 . 181 . 241 . 207 . 128 . 247 256 GREAT BARRIER REEF EXPEDITION hebes, Acropora ..... heliopora, Diploastrea ..... hemprichii, Lobophyllia .... hirsuta, Goniopora ..... horrescens, Acrohelia ..... lxorrida, Echinopora ..... humilis, Acropora ..... hystrix, Seriatopora ..... incisa, sp.n., Balanophyllia .... informis, Montipora ..... intermedia, Acropora ..... irregularis, sp. n., Alveopora .... irregularis, Anomastrea .... jeulini, sp. n., Acropora .... Icauaiensis, Madrepora .... lacera, Oxypora ...... lactuca, Tridacophyllia .... laevis, sp. n., Acropora ..... lamellina, Coeloria ..... lamellosa, Echinopora ..... lanuginosa, Porites ..... lalistella, Acropora ..... Leptastrea ....... Leptoria ....... Umax, Herpolitha ..... Lithophyllia ...... lobata, Goniopora ..... lobata, Porites ...... Lobophyllia ...... lutkeni, sp. n., Acropora .... macrostoma, Acropora .... Madrepora ...... mantonae, sp. n., Goniastrea .... mantonae, sp. n., Turbinaria . mayeri, Coeloseris ..... micranthus, Dendrophyllia .... micranthus var. grandis, var. n. Dendrophyllia microconos, Hydnophora . ' . microphthalma, Cyphastrea .... Millepora ....... millepora, sp. n., Montipora .... minor, nom. nov., Goniopora minor, Thecocyathus ..... Montipora ...... morlenseni, sp. n., Alveopora .... musica, Tubipora ..... myriophthalma, Aslraeopora Orbicella ....... often, sp. n., Acropora ..... Oulophyllia ...... Oxypora ....... PAGE Pachyseris . 163 pacijica, Acropora . 202 paeonia, Tridacophyllia . 160 palifera, Acropora . 218 patelliformis, Fungia . . 154 patula, Acropora . 215 paumotuensis, Fungia . 153 Pavona .... . 160 pectinata, Goniastrea . . 135 peltata, Turbinaria . 174 peltata var. gibiari var. n., Turbinaria . . 175 phrygia, Leptoria . 150 pistillata, Stylophora . . 107 Pocillopora . 108 Podobacia .... . 155 Porites . 237 prolifera, Montipora . . 190 prolixa, Acropora . 226 prominula, sp. n., Montipora . . 182 pruinosa, sp. n., Leptastrea . . 116 pulchra, Acropora . 203 pulchra var. alveolata, Acropora . 204 pulchra, var. stricta, Acropora . 204 punctata, Stylaraea . 236 purpurea, Leptastrea . . 115 quelchi, Acropora . 209 radians, Sympliyllia . . 146 ramosa, Montipora . 184 recta, Symphyllia . 144 retiformis, Goniastrea . . 133 robusta, Halomitra . 155 rosaria, Acropora . 224 rubrum, Flabellum . 105 savignyi, Cynarina . 137 scutaria, Fungia . 152 seplata, Stylophora . 108 serailia, Cyphastrea . 118 Seriatopora . 106 simplex, sp. n., Caulastrea . . 139 socialis, Montipora . 188 solida, Porites . 242 speciosa, Favia . . 127 speciosa, Pachyseris . 164 spectabilis, Acropora . . 220 squamosa, Acropora . 216 stelligera, Favia . 128 Stephanophyllia . . 154 stephensoni, sp. n., Porites . . 238 stephensoni, sp. n., Turbinaria . 178 Stylaraea .... . 236 Stylophora .... . 107 sulcata, Montipora . 191 suppressa, sp. n., Porites . 245 surculosa, Acropora . 214 Symphyllia . 144 PAGE 217 166 143 233 123 120 221 107 166 195 200 234 165 227 121 158 160 230 149 119 243 216 114 150 154 138 232 242 142 229 214 121 136 179 164 171 173 151 118 248 182 233 103 180 235 251 180 123 229 146 158 MADREPORAMA, HYDROCORALLIXAE, HELIOPORA AND TUBIPORA— CROSSLAXD 257 talpina, Polyphyllia tenuidens, Goniopora . tertia, sp. n., Montipora Thecocyath us torresiana , Pachyseris . tortuosa, Millepora Trachyphyllia transversa, Leptastrea . Tridacophyllia tubigera, Acropora Turbinaria undans. sp. n.. Montipora PAGE vacua, sp.n., Orbicella ..... 124 vacuum, sp. n., Flabellum .... 106 valenciennesi. Facia . . . . .126 variabilis var. pachyclados, Acropora . . 222 various, Pavona ..... 162 vaughani, sp. n., Porites ( Synaraea ) . . 247 velata, sp. n., Dendrophyllia .... 173 venosa var. angulosa, Montipora . . . 188 verrucosa, Montipora . . . . .193 verrucosa, Pocillopora . . . . .111 virens, Favites ...... 130 vitiensis, LithophyUia ..... 138 yongei, sp. n., Balanophyllia .... 167 PAGE 155 231 192 103 164 249 101 115 159 208 174 190 DESCRIPTION OF PLATE I. Fig. 1. — Flabellum vacuum, sp. n. x 1*1. B.M. Reg. No. 1934.5.14.620. Fig. 2. — Pocillopora eydouxi M. E. and H. x G.B.R.E. No. 629. Fig. 3. — Flabellum vacuum, sp. n. x 1. B.M. Reg. No. 1934.5.14.620. Fig. 4. — Leptastrea bottae M. E. and H. x 1. G.B.R.E. No. 407. Fig. 5. — Leptastrea purpurea Dana, x 1. B.M. Reg. No. 1934.5.14.420. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (Xat. Hist.). Reports. Yol. VI, Xo. 3. PLATE I. [Adlard <& Son, Ltd., Impr. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (Nat. Hist.). Reports, Vol. VI. No. 3. PLATE II. i DESCRIPTION OF PLATE II. Fig. 1. — Orbicella vacua, sp. n. x '8. G.B.R.E. No. 400. Fig. 2. — Leptastrea bottae M. E. and H. x 3‘5. B.M. Reg. No. 1934.5.14.444. Fig. 3. — Leptastrea bottae M. E. and H. x 3'8. B.M. Reg. No. 1934.5.14.444. Fig. 4. — Orbicella vacua, sp. n. x 2. G.B.R.E. No. 433. Fig. b.—Stylophora septata Gardiner, x 3. G.B.R.E. No. 369. Fig. 6.— Euvhyllia alabrescens (Chamisso). X 1*6. G.B.R.E. No. 431. DESCRIPTION OF PLATE III. Fig. 1. — Leptastrea pruinosa, sp. n., Holotype. x 4'9. B.M. Reg. No. 1934.5.14.630. Fig. 2 .—Orbicella vacua, sp. n. x 3. G.B.R.E. No. 433. Fig. 3. — Leptastrea purpurea Dana, x 3. B.M. Reg. No. 1934.5.14.420. Figs. 4-4«. — Caulastrea simplex, sp. n., Holotype. x 2. B.M. Reg. No. 1934.5.14.615. GREAT BARRIER REEF EXPEDITION 1928-29 Brit. Mus. ( Nat . Hist.). Reports, Yol. VI, No. 3. PLATE III. [Adlard & Son, Ltd., Impr. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (Nat. Hist.). Reports, Vol. VI. No. 3. PLATE TV. [Adlard & Son, Ltd., Impr. DESCRIPTION OF PLATE IV. Figs. 1 and 2. — Cynarina savignyi Briigg. x 1. B.M. Reg. No. 1934.5.14.492. Fig. 3. — Echinopora horrida Dana. X 1. G.B.R.E. No. 26. DESCRIPTION OF PLATE V. Fig. 1. — Favites aspera Verrill. x 1'3. G.B.R.E. No. 166. Fig. 2.— Favites aspera Verrill. x 1'5. G.B.R.E. No. 155. GREAT BARRIER REEF EXPEDITION 1928-29. Bnt. Mits. (Xat. Hist.). Reports. Yol. VI. No. 3. PLATE V. [Adlard & Son, Ltd., Impr. GREAT BARRIER REEF EXPEDITION 1928 29 Brit. Mas. (Xat. Hist.). Reports. Yol. VI. No. PLATE VI. [Adlard c£- Son, Ltd., Impr. DESCRIPTION OF PLATE VI. Fig. 1. — Favites virens Dana. G.B.R.E. No. 114. X 1. Fig. 2.— Favites virens Dana. G.B.R.E. No. 115. X 1. DESCRIPTION OF PLATE VII. Fig. 1.- — Goniastrea mantonae, sp. n. x 1. G.B.R.E. No. 120. K5. Fig. 2. — Goniastrea mantonae , sp. n. x 1 . G.B.R.E. No. 128. I GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (Xat. Hist.). Reports, Yol. YI, No. 3. PLATE VII. [Aillard & Son, Ltd., Impr. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (Nat. Hist.). Reports, Vol. VI, No. 3. PLATE VIII. DESCRIPTION OF PLATE VIII. Fig. 1. — Acanthastrea echinata Dana, x 2. G.B.R.E. No. 151. Fig. 2. — Goniastrea benhami Vaughan, x 1. G.B.R.E. No. 51. Fig. 3.— Acanthastrea echinata Dana, x 1-2 G.B.R.E. No. 144. DESCRIPTION OF PLATE IX. Fig. 1. — Acanthastrea echinata Dana. X 3. G.B.R.E. No. 151. Fig. 2. — Acanthastrea echinata Dana. X 3. G.B.R.E. No. 144. Fig. 3. — Lohophyllia corymbosa (Forskal). X 1. G.B.R.E. No. 447. Fig. 4. — Lithophyllia vitiensis (Briigg). x 3. Stn. XXIV. B.M. Reg. No. 1934.5.14.410c. GREAT BARRIER REEF EXPEDITION 1928-29, Brit. Mus. (Nat. Hist.). Reports. Yol. VI, No. 3. PLATE IX. [ Adlard Son , Ltd.. Impr. DESCRIPTION OF PLATE XXVI. Fig. 1. — Monti pora digitata (Dana). X '4. B.M. Reg. No. 1934.5.14.281. Fig. 2 . — Montipora fossae sp. n. x '4. B.M. Reg. No. 1934.5.14.286. Fig. 3. — Montipora ramosa Bernard. X '5. B.M. Reg. No. 1934.5.14.275. Fig. 4. — Montipora ramosa Bernard. X '5. B.M. Reg. No. 1934.5.14.273. Fig. 5. — Montipora venosa, var. angulosa, Klz. x '4. B.M. Reg. No. 1934.5.14.394. DESCRIPTION OF PLATE XXVII. Fig. 1. — Montipora erythraea Marenz. x 4. G.B.R.E. No. 136. Fig. 2. — Montipora erythraea Marenz. x 8. G.B.R.E. No. 136. Fig. 3. — Montipora ramosa Bernard, x 5. B.M. Reg. No. 1934.5.14.300. Fig. 4. — Montipora granulosa Bernard. X 5 '8. G.B.R.E. No. 113. Fig. 5.— Montipora venosa var. angulOsa Klz. x B.M. Reg. No. 1934.5.14.394. Fig. 6. — Montipora prominula, sp. n. x 5. G.B.R.E. No. 327. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (X at. Hist.). Reports, Yol. VI. No. 3. PLATE XXVII. [Acllard ■ 5 11. „ 3 miles E. . 32 . 21°52 6 17. „ ■ w »■.' ^ i,, ' 7 . 22. „ • „ . 22°42 8 . 27. „ . I.T.O. . 45 . 23°22 9 31. „ . 3 miles E. . 32 . 23°00 10 4.ix . .. . 23°45 11 6. „ . I.T.O. . 23°91 12 11. „ . 3 miles E. 32 24°13 19. „ R.F. 13 20. „ . 3 miles E. . ' . 24°53 14 26. „ • »» • . 25°11 28. „ R.F. 15 2.x . 3 miles E. . . 25°10 16 3.,, 17 18 Cn 00 . 4e..- ■ . 25°40 . 25°93 19 . 20. „ . O.T.O. . 225 . 26°05 20 • » » >600 v P- [ * 21 . 22.,, . 3 miles E. . 32 . 22 • 23.,, » . 26°50 23 . 2.xi • •> • j . 27°24 24 6. „ . 27°33 Net. S e s c s c e s e F S c s c s c s c s c s c N s c c s c F S c s 0 F m. wire 0 10 20 30 40 25 S S C F . S 180-0 N „ N „ . N 250- 0 nr. S c F S c . s 0 s c s a o e 5 1 - - - - 2 - — - - - 1 1 - - - - 11 - - - - - 5 4 1 4 2 9 12 9 39 29 — — — — — — 1 — — — — — — — — — — — — 4 -1--- — - - — — — — - — — — — — -10 ------ 2---- — - -- -- - -- 4 ------1------------3 -------- 1 — — - -- -1---7 -------- -------2---6 2 - - - - 2 - - - 2 1 2 1 1 1 o 12 1 ----11 8 - - - - 4 - - - - 1 - - - - 1 o 1 2 ------1---1----3---2 --- - I-?---- - i 1-5 1--23 — — — — — 211 3 - 1 - 1 - 1 1 1 1 - 3 — — — — 1 — 1 — — — — - — -52 3 2 1 2 - - 8 ----1-1-1 ---4S---9 ---1-12-1 - 5 - 5 3 17 1-------- 2 7 5 1 1 2 - - - 2 2 - - 1 1-------4 — — 1 — — — — 116 ------- - 10 _ _ 1 ----- 9 — - -1---11- - 18 1------37 — — — 5 — — — — 2 — — 341 — — 2 — — _ — — 12 — - -- -- 131 — — — I® Aglaura HYDROMEDUSAE — KRAMP 321 Table XVIII.- -List of Stations where Hydromedusae were Collected, with Number of Specimens in the Preserved Collection. — -Continued. .tation Date. Position. Depth Temp. lumber. 1928. (m.) (5C.) 25 . 16. xi . 3 miles E. . 32 . 26°57 26 . 19. „ . I.T.O. 57 . 26°61 30 30a 31 21. „ 23. „ 25. „ 28. 29. „ 2.xii 5. 33 . 13. , • .. . 28° 17 . 8 C _____ 1 3 - - 6 - - - 20 - 11 1 19 - 4 - - 3 - 1 1 - 1 1 2 34 19. , • .. . 28°42 . 8 C - - - 1 3 - - - - 29 - 3 17 3 32 - 4 - _ 2 - 1 2 2 1 1 2 8 1 27 7 35 . 27. , . . 29°43 . s - 2 - 3 - - 17 - - - 8 44 12 1 - 25 4 3 - 7 63 7 1929. 4.i 14. „ 21. „ 30. „ 13. ii 18. „ 26.,, 27.. , 28.. , 3 milps E. O.T.O. L.O.R.F. 3 miles E. C.B. LI. I. O.C.P. 32 >600 32 26a90 26°91 27°34 31 32 31 32 30 31 >600 28°56 28°47 28°94 28°89 28t>64 29°00 28°81 27°97 ;6 .. .. I.C.P. . 33 27°61 s c '-7 4.iii 3 miles E. . 31 28°42 s c 8 15. „ .. 30 28°27 s c 9 • U. „ I.P.P. . 46 28°23 s c 0 • 18. .. O.P.P. . >400 28°24 S 8 C 400-0 170-0 150 0 1 • 25. „ 3 miles E. . 33 27°68 8 C • 6.1y » 32 27°62 s Net. S C F N 50-0 S C F S 600-0 C 580-0 s 8 C B.S. S 500-0 S' 1 - - 3 - - 2 — - — - 09 - - 1 14 - - - 3 - - - 1 3 11 - - 4 2 — — 1 10 - 6 2 1 1 - 1 1 - - - - 10 - 1 4 6 - 1 - - - - 3 1 - - - - 3 1 1 1 — — — 1 - - — 11 11 8 1 7 2 1 - - 6 - - 1 - - 3 - - 13 - 25 - 11 113 23 - 3 - 4 - - 2 - 5 1 1 - 1 - 3 - 1 1 6 10 ---3- -3-10- --29- - - 1 14 - - - 1 13 - 2 - 18 30 4 4 - 10 - 1 - - 4 - 3 9 39 4 10 3 20 3 11 9 33 - 1 - 2 31 - 1 16 1 1 1 16 3 125 51 20 8 1 36 13 16 1 16 8 - - - 2 11 2 4 - 13 1 -13- 1 4 - 7 1 1 - - 13 7 2 24 - 1 - 1 1 5 13 2 4 - 3 - 1 1 1 1 24 - 2 - 14 4 - - 3 15 - 2 3 1 1 1 - - - 6 - 1 - 15 10 10 1 3 - 8 4 - Cunina octonaria. 322 GREAT BARRIER REEF EXPEDITION Table XYIII. — List of Stations where Hydromedusae vjere Collected, with Number of Specimens in the Preserved Collection. — Continued. Station Date. Position. Depth Temp. Net. §4 « js Si ©i 3 © ©3 g. © © 3 .3 number. (m.) (°C.) ft? o ' < H ft* Wi 53 . 13. iv . 3 miles E. . 32 . 26°90 . S 0 54 20. „ 31 . 25°96 . s ■ ■ 2 - - - - - - e? a; 55 . 26. „ . JJ „ 56 7 . v 19 >>' 57 . 18. „ 11 32 58 . 25. „ 11 ,, 59 . 31. „ . 91 •> 60 7 . vi . 11 61 . 14. ,, . 11 31 63 • 24. „ . 11 32 66 . ll.vii . 11 30 67 • 17. „ . „ 32 26°52 25°27 24°68 24°09 23°63 23°99 22°55 22°10 21°41 21°27 17 12 2 112 2 - - - 1 - - - - 4 - 3 - - - - - - 4 - - 3 1 - - - - 2 - 1 1 1 - - 1 2 - - 1 1 - - - - 1 - 3 - 3 . g. 13 20 8 19 4 48 13 11 1 12 15 18 13 25 8 15 12 15 15 13 7 Total number of specimens ---------------3-- - 13 10 33 81 18 13 19 335 10 22 9 290 255 234 54 45 157 20 11 27 891 834 11! S : stramin net. 3 miles E.: 3 miles E. of Low Isles. O.C.P. : outside Cook’s Passage. C-- coarse silk net. 2 miles N.E. : 2 miles N.E. of Low Isles. I.C.P. : inside Cook’s Passage. F : fine silk net. I.T.O. : in Trinity Opening. I.P.P. : inside Papuan Pass. N : Nansen net, vertical. O.T.O. : outside Trinity Opening. O.P.P. : outside Papuan Pass. B.S. : bottom stramin net. Li. I. : L.O.R.F. : C.B. : off Lizard Island, lagoon over reef flat, oft Cape Bedford. R.F. : reef flat, entrance to anchorage, Low Isl SUL IBLid EXPLANATION OF PLATES. PLATE I. Fig. 1. — Octotiara russelli n. sp. Lateral view of the medusa. X 7. Fig. 2. — Octotiara russelli n. sp. Part of umbrella margin. X 25. Fig. 3. — Octotiara russelli n. sp. Tentacle bulb, cc circular canal, rc radial canal. X 40. Fig. 4. — Phialucium condensum n. sp. Oral view of the medusa. X 15. PLATE II. - — - Fig. 5. — Eirene kambara. Oral view of the medusa, x 14. Fig. 6. — Eirene menoni n. sp. Lateral view of the medusa. X 6. Fig. 7. — Stomach of Liriope tetraphylla with medusiferous stolons of Cunina. X 35. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (Nat. Hist.). Reports, Yol. VI, No. 4. rc PLATE I. [ Adlard & Son, Ltd., I mpr. GREAT BARRIER REEF EXPEDITION 1928-29 Snf. Mus. (Nat. Hist.). Reports, Yol. VI, No. 4. PLATE II. \Adlurd & Son, Ltd., Impr. -55 BRITISH MUSEUM (NATURAL HISTORY) GREAT BARRIER REEF EX 1928-29 SCIENTIFIC REPORTS VOLUME VI. No. 5 THE SERGESTIDAE OF THE GREAT BARRIER REEF EXPEDITION BY ISABELLA GORDON, D.Sc.. Ph.I). WITH SIX TEXT-FIGURES LONDON : PRINTED BY ORDER OF THE TRUSTEES OF THE BRITISH MUSEUM SOLD BY B. Quaritch, Ltd., 11 Grafton Stukhj', New Bond Stb.tst, London, VV. I ; Wheldon & Wesi.ey, Ltd., 83/84 Bekwick. Stbeet, London, W. I H.M. Stationery Omci, London, S.VV. I AND AT The British Moskcsi (Natckal History}, Cbohwbli. Road, London, S.W.7 1958 [H//! right! revived] • Pries Five Shillings [Issued 31st January, 195ti\ • * / V* *y. g - >: J A *• i VV7 ^ j ^ / ^<^4 L H ^ THE SERGESTIDAE OF THE GREAT BARRIER REEF EXPEDITION BY ISABELLA GORDON, D.Sc., Ph.D. SYNOPSIS. The paper gives the occurrence of two species of the genus Lucifer in the Great Barrier Reef area during the year July 1928-July 1929. L. penicillifer Hansen is by far the commoner species ; it occurred with fair regularity throughout the year, the month of September excepted. Spermatopliores were present, in the distal portion of one vas deferens only, practically throughout the year, suggesting that there is no fixed breeding period. L. typus H. M.-Edw. occurred in small numbers between the end of July and the end of November 1928 but the two species were seldom present at the same time. INTRODUCTION The Sergestidae of the Great Barrier Reef Expedition all belong to the subfamily Luciferinae which comprises the single aberrant genus Lucifer Y. Thompson (= Leucifer H. Milne-Edwards). This genus was revised by Hansen (1919, pp. 48-65, pis. iv and v) who reduced the number of known species to three, adding that “ all the remaining names in the literature must be cancelled for ever either as synonyms or as quite unrecognizable ” (p. 50). In addition, he described three new species from the “ Siboga ” material. These- six species fall into two groups, one with long eye-stalks comprising L. typus IT. M.-Edwards. and L. orientalis Hansen, the other with short eye-stalks including the remaining four species — L. faxoni Borradaile, L. Jianseni Nobili. L. intermedins Hansen and L. penicillifer Hansen. Hansen also gives keys to the determination of the species in each group and good figures of the petasma of each male (1919, pp. 52-57 and pis. iv and v.). Only two species are represented in the abundant material of the Great Barrier Reef Expedition, namely Z. typus H. Milne-Edwards and L. penicillifer Hansen — in each the petasma agrees with Hansen's illustrations. The latter species is by far the more plentiful,, over 1,350 specimens having been obtained at 48 Stations as against 67 of L. typus from 15 Stations. L. penicillifer occurred with fair regularity throughout the year from 27th July, 1928, to 1 7th July, 1929, except for the period between 24th August and 2nd October, 1928. Z. typus , on the other hand, was present during the weeks when L. penicillifer seemed to be absent and, although it was obtained in small numbers between 30th July and 29th November, 1928, the two species were rarely present in the same hauls of the townets (Stations 2, 8. 15 and 21). Dakin and Colefax (1940, pp. 149-150) record only two species from the coastal waters of New South Wales, namely Z. typus and Z. Jianseni Borradaile. Here the former species is stated to be rare, the latter being much more common. The pair of tiny spinules near the distal end of the ventral margin of the last abdominal somite in the female of Z. Jianseni figured by Dakin in fig. 24 1« is present also in Z. typus and in Z. penicillifer . vi. 5. 22 '324 GREAT BARRIER REEF EXPEDITION Note — The males frequently have a spermatophore in the distal portion of either the left or the right vas deferens ; to avoid repetition these males are listed under each Station as £‘ adult males.” Special mention is made only of specimens in which the spermatophore was protruding from the genital opening of the male, or from the thelycum of the female. Family Sergestidae. Subfamily Luctferinae . Genus Lucifer Vaughan Thompson. Lucifer, J. Vaughan Thompson, 1829, p. 58, pi. vii, fig. 2. Lucifer, H. J. Hansen, 1919, p. 48. Leucifer, H. Milne-Edwards, 1837, p. 467 ; Barnard, 1950, p. 644. Group A. — Eye-stalk Long. Lucifer typus H. Milne-Edwards. Leucifer typus H. Milne-Edwards, 1837, p. 469. Lucifer typus, Hansen, 1919, p. 53, pi. iv, figs. 6a-6& ; also references and synonymy. ( non Leucifer typus Stebbing 1914, p. 28 which = L. penicillifer). Lucifer typus, Edmondson, 1925, p. 5 ; Cecchini, 1928, p. 52 ; Zariquiey, 1946, p. 58, fig. 58 ; Hiatt, 1947, p. 241-2. Occurrence : St. 2, 30. vii. 28. 3 miles east of Low Islands. I metre stramin net. 1 adult and 1 immature d (with L. penicillifer). St. 8, 24.viii.28. 16° 30' S., 145° 52' E. (Trinity Opening). I metre stramin net. 1 $ (with L. penicillifer). St. 9, 31 . viii.28. 3 miles east of Low Islands. 1 metre stramin net. 1 d- St. 11, 6.ix.28. 16° 24' S., 145° 52' E. (Trinity opening). 1 metre stramin net. 2 $ and one incomplete £ minus head region. Coarse silk townet. 1 $. St. 12, ll.ix.28. 3 miles east of Low Islands. 1 metre stramin net. 1 ?, 3 adult d (one has the spermatophore projecting from the genital duct) and 3 immature d- St. 13, 20. ix. 28. 3 miles east of Low Islands. 1 metre stramin net 2 $ and 6 young, 1 adult d St. 13. Fine Silk International net- 1 $. ,, Coarse Silk International net. 2 immature d- St. 14, 26. ix. 28. 3 miles east of Low Islands. 1 metre stramin net. 1 im- mature o, ,, Coarse silk townet. 1 2, 1 immature d- THE SERGESTIDAE OF THE EXPEDITION— GORDON 325 St. 15, 2.x. 28. 3 miles east of Low Islands. Coarse silk townet, 1$ (with L. penicillifer). ,, 1 metre st.ra.min net. 1 $. St. 16, 3.x. 28. 3 miles east of Low Islands. Closing coarse silk townet. Surface. Horizontal haul. 2 $. ,, n. 10 m. (Vert, dist.) 1 $. ,, III. 20 m. (Vert, dist.) 1 immature St. 18, 15.x. 28. 3 miles east of Low Islands. 1 metre stramin net. 2 $. ,, Coarse silk townet. 1 $. ,, Fine silk townet. 2 $, 1 young. St. 19, 20.x. 28. 16 20' S., 146' 3' E. (outskle Trinity Opening). 1 metre stramin net. 1 immature $. St. 20, 20.x. 28. 16' 19' S., 146° 7' E. (outside Trinity Opening). Coarse silk townet. 1 $. St. 21, 22.x. 28. 3 miles east of Low Islands. Fine silk townet. 1 $ (with L. penicillifer). St. 28, 23.xi.28. 16 19' S., 146° 5' E. (outside Trinity Opening). 1 metre stramin net. 5 $, 3 adult and ] immature ,, Coarse silk townet. Tube broken — 1 $ and 1 immature G adhering to sides of tube. St. 29, 29.xi.28. 16° 17' S., 146° 2' E. (outside Trinity Opening). Bottom stramin net. 6 $ and young (one $ with empty spermatophore case projecting from the thelycum) ; 3 adult and 2 immature <$. Recorded Distribution. — This species is very common in the warmer parts of the Atlantic Ocean — southern parts of the Gulf Stream, Sargasso Sea, North and South Equa- torial Currents and Guinea Current— see map given under the specific name L. reynaudi by Ortmann (1893, Tafel x). The most southern Atlantic records are 28° 43' S., 25° 14' W. and 40° 32' S., 52° 2' W. It is rather rare in the Bay of Bengal and in the Pacific Ocean from the Great Barrier Reef area to Manilla. L. orientalis Hansen is much commoner in the Pacific and Indian Oceans but writers prior to Hansen (1919) did not realize that these two species are distinct and many of the earlier records of L. typus or its synonyms refer to L. orientalis. Remarks.— Hansen, 1919, p. 52, states that females of L. typus are not easily distin- guishable from those of L. orientalis. All the males in the present collection agree with Hansen’s figures and description of L. typus as regards the telson and the petasma, which possesses a long-stalked hook, so it is reasonable to suppose that the females belong to the same species. The telson of the male is represented, in dorsal and in lateral aspect, in figs. 2 and 3 respectively. The large ventral cushion (which is lacking in the female and in immature males) is situated some distance from the apex ; beyond the distal pair of lateral spines the telson narrows rather abruptly and the apex bears 3 pairs of spines. The longest pair are shorter than the width of the apical margin of the telson and each has 4 spinules on the proximal half of the inner side. Burkenroad (1934, pp. 133-134) found a great preponderance of males in his material of L.faxoni Borradaile ; he thought that the females probably mate with two males since only one spermatophore is formed at a time. Although the present samples of L. typus are vi. 5. 22§ 326 GREAT BARRIER REEE EXPEDITION small, the total number of specimens being only 67, the sexes appear to be almost equally well represented — if anything the females are slightly in excess of the males. Text-fig. 1 represents part of the thorax and first abdominal somite of a male from St. 29, which has 0-5 mm. E E m 6 Text-pigs. 1-3. — Lucifer typusf H. M.-Edw.). (1) Posterior part of thorax and first abdominal somite of a male from St. 29, in lateral aspect, showing fully formed spermatophore in distal part of right vas deferens and two developing spermatophores in abdominal parts of the vasa deferentia. p., Petasma ; a.m., appendix masculina ; III, base of third pereiopod. (2) Telson of male, in dorsal aspect, with apex more highly magnified. (2a) Apex of telson of another specimen, highly magnified. (3) Telson of male, in lateral aspect. been stained with lignin pink and mounted in polyvinyl-lacto-phenol. The contents of the spermatophore reacted in three ways to the stain, becoming dark red in the stalk or neck, pinkish-white in the distal two thirds of the flask and rather brownish in between. In the abdominal region of the vasa deferentia the contents of two other spermatophores are being assembled. The oval areas indicated by broken lines, the “ gelbe Secretmasse ’ THE SERGESTIDAE OF THE EXPEDITIOX— GORDON 327 of Rosenstadt (1896, Taf. 33, fig. 41) stain pinkish- white like the distal part of the flask ; the larger developing spermatophore seems to belong to the left, the smaller to the right, vas deferens, suggesting that the two sides function alternately. In Text-fig. 1 the position of the petasma and of the appendix masculina on pleopod 2 is shown ; Bargmann (1937, p. 346) thinks that in Euphausia superba Dana the spermatophore is received by the modified portions of the second pair of pleopods and passed to the first pair of pleopods (petasma) during copulation. When males are found with the spermatophore protruding from the vas deferens, as they sometimes are, this is doubtless due to the shock received by fully ripe specimens at the time of fixation. Many years ago I figured what I called spermatophores in EuryrJiynchits ivrzesnioivshii Miers, but these are undoubtedly artefacts ; the males must have been very ripe at the time of fixation and the shock of sudden immer- sion in the fixative must have caused emission of the seminal fluid, which hardened into irregular lumps (Gordon, 1935, p. 333, Fig. 18a-e). I know of no species of the Caridea having spermatophores of the kinds usually found in Euphausiacea and in the Peneidea, to which the Sergestidae belong. Unfortunately this mis-statement of mine has been repeated by Dr. Balss in his recent monumental work, now nearing completion, namely the Decapoda in Bronn’s Klassen und Ordnungen des Tierreichs (Balss, 1944, Lief. 5, p. 619). Group B. — Eye- stalk Short. Lucifer penicillifer Hansen. Lucifer penicillifer Hansen, 1919, p. 59, pi. v, figs. 2 a-k. Leucifer penicillifer, Barnard, 1950, p. 645, fig. 121. Leucifer typus Stebbing, 1914, p. 28 — see Barnard, 1947, p. 384. Occurrence : St. 1, 27.vii.28. 3 miles east of Low Islands. 1 metre stramin net. About 30 specimens, 2, d1 and young. „ Coarse silk net. 8 $ and young, 5 $. St. 2, 30.vii.28. 3 miles east of Low Islands. 1 metre stramin net. 10 $, 6 adult d1, 6 young (with L. typus). ,, Coarse silk net. 1 $, 4 adult with L. typus. St. 21, 22.x. 28. 3 miles East of Low Islands. 1 metre stramin net. 2 $, 2 adult d (part of stalk of spermatophore projecting from duct of one d, the rest may have been broken off). ,, Coarse silk townet. 2 $. ,, Fine silk townet. 1 $, with L. typus. St. 23, 2.xi.28. 3 miles east of Low Islands. Coarse silk townet. 2 $. St. 24, 6.xi.28. 3 miles east of Low Islands. Coarse silk townet. 1 slightly immature $, 1 d- St. 25, 16.xi.28. 16° 19' S., 146° 5' E. (outside Trinity Opening). Coarse silk townet. (Tube broken.) 1 immature ?, 1 d adhering to plug of cotton wool. ,, 1 metre stramin net. (Tube broken, specimens missing) ? St. 27, 21 .xi.28. 3 miles east of Low Islands. 1 metre stramin net. 1 $. St. 30, 28 . xi . 28. 3 miles east of Low Islands (outside Trinity Opening). 1 metre stramin net. 5 $, 2 young d- St. 30a, 29. xi.28. Coarse silk townet. 5 $, 3 adult d- St. 31, 3.xii.28 (or 2.xii). Bottom stramin net. About 30 specimens, $, d and young. St. 32, 5.xii.28. 3 miles east of Low Islands. 1 metre stramin net. 1 $. St. 33, 14.xii.28. 3 miles east of Low Islands. 1 metre stramin net. 2 $ (one slightly immature), 1 adult and 2 young d- St. 34, 19.xii.28. 3 miles east of Low Islands. 1 metre stramin net. 10 speci- mens, $, d and young. „ Coarse silk net. 16 $ and young, 3 d- St. 35, 27.xii.28. 3 miles east of Low Islands. 1 metre stramin net. About 35 specimens, $, d and young ; a few of the males are adult. ,, Coarse silk townet. 9 $ and young, 5 d- St. 36, 4.i.29. 3 miles east of Low Islands. 1 metre stramin net. 15 $ and . young, 5 d- „ Coarse silk net. 10 $ and young, 5 immature d- St. 37, 14 . i . 29. 3 miles east of Low Islands. 1 metre stramin net. 30 speci- mens, mostly $ and young, a few rather immature d- „ Coarse silk townet. 20 specimens, d (a few adult), $ and young. St. 38, 21.L29. 3 miles east of Low Islands. 1 metre stramin net. 15 $ and young, 5 d- St. 39, 30.i.29. 3 miles east of Low Islands. 1 metre stramin net. About 50 specimens, $ and young, and d (some adult ; one with spermatophore projecting from duct). ,, Coarse silk townet. 16 $ and young, 2 adult and 2 immature d- St. 40, 6.ii.29. 3 miles east of Low Islands. 1 metre stramin net. 14 $ and young, 8 d (mostly adult). ,, Coarse silk townet. About 50 specimens, mostly $ and young (one $ has a spermatophore projecting from the thelycum) ; 8 are d, but not all are fully grown. THE SERGESTIDAE OF THE EXPEDITION— GORDON 329 St. 42, 18.ii.29. 3 miles east of Low Islands. 1 metre stramin net. 6 9 and young, 3 d- „ Coarse silk townet. 22 specimens, $, d and young. St. 43, 26.ii.29. 15° 16' S., 145° 26' 30" E. (off Cape Bedford). 1 metre stramin net. 22 9 and young, 8 d — some rather young. ,, Coarse silk townet. 10 $ and young, 8 d, mostly adult. St. 44, 27.ii.29. 14° 44' S., ]4 5C 27' 30" E. (oft* Lizard Island). 1 metre stramin net. 10 $, 10 d — most of the specimens rather immature. „ Coarse silk townet. 30 specimens, $, young, and rather immature d- St. 45, 28.ii.29. 14" 31' S., 145° 35' E. (outside Cook’s Passage). 1 metre stramin net. Vertical haul, 500 metres of wire out. 8 9 and young, 2 young d- ,, Coarse silk townet. 4 $ and young. St. 46, 28.ii.29. 14° 31' S., 145° 35' E. (outside Cook's Passage). 1 metre stramin net. 28 specimens, mostly 9 and young, about 8 are d- ,, Coarse silk townet. 35 specimens, $, young and adult d (one d has the spermatophore projecting from the genital duct). St. 47, 4 . iii . 29. 3 miles east of Low Islands. 1 metre stramin net. 11 $ and young, 4 d (one with spermatophore projecting from duct.) „ Coarse silk townet. 2 2 d- St. 48, 15. iii. 29. 3 miles east of Low Islands. Coarse silk townet. 2 1 immature d> 3 young. ,, 1 metre stramin net. 22 $ phree have each one spermatophore projecting from the thelycum), 17 adult and young d, 15 young specimens. St. 49, 17. iii. 29. 15° 47' S., 145° 41' E. (inside Papuan Pass). 1 metre stramin net. 20 specimens, 9, d and young. ,, Coarse silk townet. 15 9 and young, 5 d- St. 50, 18. iii. 29. Outside Papuan Pass. 1 metre stramin net. 2 9, 2 adult and 1 immature d- St. 51, 25. iii. 29. 3 miles east of Low Islands. 1 metre stramin net. 15 9 and young, 7 d some immature. ,, Coarse silk townet. 4 9 and young. St. 52, 6.iv.29. 3 miles east of Low Islands. 1 metre stramin net. 8 9 and young, 2 adult and 2 immature d- ., Coarse silk townet. 9 9 and young (one 9 with an empty spermatophore case projecting from the thelycum), 1 d- St. 53, 13.iv.29. 3 miles east of Low Islands. 1 metre stramin net. 9 9 and young, 2 adult and 2 rather immature d- ,, Coarse silk townet. 15 9 and young, 11 d some of which are not fully grown. St. 54, 20.iv.29. 3 miles east of Lowr Islands. 1 metre stramin net. 27 9 and young, 2 adult and 1 immature d- ,, Coarse silk townet. 16 9 and young (one 9 with spermatophore projecting from thelycum), 4 d- St. 55, 26.iv.29. 3 miles east of Low Islands. 1 meter stramin net. 30 speci- mens, mostly 9 and young, a few immature d- 330 GREAT BARRIER REEF EXPEDITION St. 55, Coarse silk townet. 30 ? and young, 7 adult and 6 immature <$. St. 56, 7.V.29. 3 miles east of Low Islands. 1 metre stramin net. 3 $, 11 young, 7 adult $ (the spermatophore projecting from the duct in two $). ,, Coarse silk townet. 18 $ and young, 8 <$. St. 57, 18. v. 29. 3 miles east of Low Islands. 1 metre stramin net. 20 speci- mens— adult $, and <$, and young (one d has the spermatophore projecting from the duct). „ Coarse silk townet. 2$, 4 adult d (one d with the spermatophore projecting from the duct). St. 58, 25.V.29. 3 miles east of Low Islands. Coarse silk townet. 5 $ and young, 5 adult d (2 each with a projecting spermatophore), 1 immature d- St. 59, 31.V.29. 3 miles east of Low Islands. 1 metre stramin net. 12 $ (one with two large ova attached to right pereiopod III, the other ova presum- ably had been shaken off), 6 adult d, 3 young. ,, Coarse silk townet. 17 specimens, $, adult d and young. St. 60, 7.vi.29. 3 miles east of Low Islands. 1 metre stramin net. 11 $ and young, 3 adult d (one with the spermatophore projecting from duct), 1 immature d- ,, Coarse silk townet. 5 $, 5 d- St. 61, 14.vi.29. 3 miles east of Low Islands. I metre stramin net. young, 8 adult and 4 immature d- ,, Coarse silk townet. 8 $ and young, 4 adult and 2 immature d- St. 63, 24.vi.29. 3 miles east of Low Islands. 1 metre stramin net. young, 7 adult and 2 immature d- ,, Coarse silk townet. 8 $ and young, 6 adult and 2 immature d St. 66, ll.vii.29. 3 miles east of Low Islands. 1 metre stramin net. young, 2 adult, 2 immature d- ,, Coarse silk townet. 5 $, 8 adult and 1 immature d- St. 67, 17.vii.29. 3 miles east of Low Islands. 1 metre stramin net. ,, Coarse silk townet. 7 $ and young, 5 d- Recorded Distribution. — Common in the Dutch East Indies area explored by the “ Siboga ” Expedition, and in the Great Barrier Reef area. Other localities mentioned by Hansen are Bay of Bengal, China Sea, Formosa Strait, Manilla and Gulf of Yedo. Remarks. — The telson of the male is represented in dorsal aspect in Text-fig. 5 and in lateral aspect in Text-fig. 6. The protuberance or cushion on the ventral surface is much smaller than that of L. typus and is situated further from the apex of the telson. The distal pair of lateral spines are situated nearer to the apex, are larger and beset with spinules on the proximal half of either side (cf. Text-figs. 2 and 3 with 5 and 6). The median pair of apical spines is poorly developed but the outer pair are equal to, or rather longer than, the width of the apical margin and have 5 to 8 spinules on each side. Owing to the small size and extreme lateral compression of the specimens it is by no means easy to detect the genital openings in either sex. Early writers thought that, as an adaptation to the compression of the body, there was a median unpaired opening in each sex. Brooks, however, corrected this in his classic study of the morphology and develop- ment of Lucifer (Brooks, 1882, p. 58) in the case of the male in which he found that the vasa deferentia are equally well developed on both sides. In the female he thought that 12 $ and 11 $ and 5 $ and 5 ?, 5 . Nat. Mus. Bull. 161, pt. 1, p. 32, pi. 8, figs. 2a, b. ] Massilina pacific ien sis Cushman, 1924, Publ. 342, Carnegie Inst. Washington, p. 66, pi. 24, figs, 1, 2. Pseudomassilina australis (Cushman). Lacroix, 1938, p. 3. The specimens referred here are rather variable both in shape and decoration. Many are unornamented as in ( Jushman’s figure of M. australis (loc. cit.) , others have the crenu- lations of M. prunfxnensis, with, intermediate forms connecting the two. Cushman does not mention the obvious relationship between these two species in his description of the later one, M. australis, though lie describes the later chambers of that species as “ wrinkled It seems possible that the two forms are the same, in which case M. australis would have to be placed in the synonymy of M. pari Jiciensis. This can only be decided by examination of the types, and the question is therefore left open, the present specimens being referred to M. australis as the majority agree with the figure of this species. Loeblich and Tappan (1953, p. 42) have included this species in the new genus Pateoris. It is however the genotype of Pseudomassilina Lacroix, 1938, and this name must have taken precedence, had the species been congeneric with the genotype of Pateoris , P. hcmerinoides Loeblich and Tappan. However, the peculiar pitted and canaliculate shell- structure of Pseudomassilina is not described as a characteristic of Pateoris, and this genus is apparently distinct from the tropical massilines separated by Lacroix. Genus Spiroloculina d’Orbigny, 1826. (99) Spiroloculina angulata Cushman. Spiroloculina grata Terquem, var. angulata Cushman, 1910 etc. (1917), U.S. Nat. Mus. Bull. 71, pt. 6, p. 36, pi. 7, fig. 5. 8. angulata Cushman. Cushman and Todd, 1944, p. 50, pi. 7, figs. 18-22. This costate species with ridged periphery is common in the samples of groups 1 , 2 and 3. It has a very wide distribution in the tropical Indo-Pacific, and has been recorded from as far south as Sydney, N.S.YV . (100) Spiroloctdina aperta Cushman and Todd. Spiroloculina aperta Cushman and Todd, 1944, Sp. Publ. No. 11, Cushman Lab. Foram. Research, p. 66, pi. 9, figs. 21, 22. This delicate species with concave periphery, 2 sharp keels and a produced apertural neck is fairly common in the samples of groups 2 and 3. 364 GREAT BARRIER REEF EXPEDITION (101) Spiroloculina communis Cushman and Todd. Spiroloculina communis Cushman and Todd, 1944, Sp. Publ. No. 11, Cushman Lab. Foram. Research, p. 63, pi. 9, figs. 4, 5, 7, 8. This strongly-built, excavated species is one of the commonest in the samples of groups 1, 2 and 3. (102) Spiroloculina clisparilis Terquem. Spiroloculina clisparilis Terquem, 1878, Mem. soc. geol. France, ser. 3, 1, p. 55, pi. 5 (10), fig. 12 ; Cushman and Todd, 1944, p. 35, pi. 5, figs. 22-31 (synonymy). Occurs sparsely in the samples of groups 3 and 4. This species has a wide warm-water distribution, and is common in the Miocene sediments of Victoria. (103) Spiroloculina foveolata Egger. Spiroloculina foveolata Egger, 1893, Abhandl. Kon. bay. Akad.Wiss. Munchen, cl. ii, 18, p. 274, pi. 1, figs. 33, 34 ; Cushman and Todd, 1944, p. 48, pi. 7, figs. 7-12 (synonymy). Fine specimens of this reticulate species were not uncommon in the samples of groups 1, 2 and 3. It has a wide Indo-Pacific distribution. (104) Spiroloculina lucida Cushman and Todd. Spiroloculina lucida Cushman and Todd, 1944, Sp. Publ. No. 11, Cushman Lab. Foram. Research, p. 70, pi. 9, figs. 30, 31. Rare in sample 10 (2 specimens). Widely distributed in the Indo-Pacific area. (105) Spiroloculina rugosa curvatura Cushman and Todd. Spiroloculina rugosa Cushman and Todd, var. curaiiwraCushman and Todd, 1944, Sp. Publ. No. 11, Cushman Lab. Foram. Research, p. 66, pi. 9, figs. 18-20. Occasional specimens of this subspecies were found in the samples of group 3. It was described from Samoa and is recorded also from the Red Sea and Sydney Harbour (N.S.W.). (106) Spiroloculina scita Cushman and Todd. Spiroloculina antillarum Parr, 1932 (not of d’Orbigny), p. 9, pi. 1, fig. 11. S. scita Cushman and Todd, 1944, Sp. Publ. No. 11, Cushman Lab. Foram. Research, p. 60, pi. 8, figs. 20, 21. Common in practically all the shallow- water samples. The specimens are generally not so large as those from southerrl Australian waters, but are otherwise typical. Genus Sigmoilina Schlumberger, 1887. (107) Sigmoilina schlumberger i Silvestri. Planispirina celata Brady, 1884 (not of Costa), p. 197, pi. viii, figs. 1-4. Sigmoilina schlumbergeri Silvestri, 1904, Mem. Pont. Accad. Nuovi Lincei, 22, p. 267 ; Cushman 1946, p. 36, pi. 6, figs. 1-4 (synonymy). Three specimens from sample 45 had the outward appearance of this species, including the very small aperture with a rudimentary tooth, giving it a crescentic appearance. On FOE A MJ X IFER A— CO L LIN S 365 sectioning one, it proved to be a typical Sigmoilina. Records of this species are mostly from the Atlantic, though Brady states that the form recorded by him is found in all the great oceans, and it is recorded by Le Roy from the late Tertiary of Siboeret Island, off Sumatra. f (108) Sigmoilina australis (Parr.) Quinqueloculina australis Parr, 1932, Proc. Roy. Soc. Victoria, 44 (x.s.), pt. 1, p.6, pi. 1, figs, la, b. Sigmoilina australis (Parr). Parr, 1945, p. 197. Three typical examples were found hi sample 45. It is common in shallow water on the Victorian coast. Genus Spirosigmoilina Parr, 1942. (109) Spirosigmoilina bradyi n. sp. Spiroloculina crenata Brady, 1884 (not of Karrer), p. 156, pi. x, figs. 24-26. Massilina crenata (Karrer). Cushman, 1910 etc. (1917) (not S. crenata Karrer), p. 57, pi. 20, fig. 2. Specimens of the form figured by Brady (loc. cit.) were frequent in the samples of group 3. A close examination of the early chambers reveals that they are sigmoiline rather than quinqueloculine, as has been previously stated. There is some variation in the degree to which the later massiline chambers embrace the earlier-formed portion, some specimens having the sigmoiline chambers obscured. Brady’s fig. 26, however, shows this character clearly. The figured specimen is one in which the crenulate development is not taken on so early as in other specimens from the same material, and hence exhibits the earlier chambers better. As this form has the characteristic growth-plan of the genus Spirosigmoilina Parr, described from the Miocene of Victoria, it is accordingly removed to that genus. Cushman and Todd (1944. p. 75) have examined the types of Karrer’s Spiroloculina crenata, and state that it is a species of Hauerina. A new specific name is therefore required for this recent species. (110) Spirosigmoilina parri n. sp. (Plate III, figs. 2a, b ; 3.) Test porcellaneous, subtranslucent, subcircular, smooth, complanate except for the raised fusiform initial portion, consisting of a proloculus and about 8 tubular chambers arranged in a sigmoid curve, followed by 3 or 4 compressed complanate spiroloculine chambers. Aperture simple, compressed, edentate. Holotype from sample 45. Dimensions of holotype : Length 0-33 mm., breadth 0-28 mm., thickness 0-06 mm. This species resembles S. tateana (Howchin) to some extent, but differs from the Miocene species in its much smaller size and more compressed and spreading later chambers. It is named in honour of my old friend the late W. J. Parr, the author of the genus. The species occurs rarely in the samples of groups 3 and 4. Genus Articulina d’Orbigny, 1826. (111) Articulina pacifica Cushman. Articulina pacifica Cushman, 1944, Sp. Publ. No. 10, Cushman Lab. Foram. Research, p. 17, pi. 4, figs., 14-18. Rare in the samples of groups 2 and 3. It has a wide Indo-Pacific distribution. 366 GREAT BARRIER REEF EXPEDITION (112) Articulina tricarinata n. sp. (Plate III, figs. 5a, b ; 6.) Test minute, translucent, surface somewhat rough, early chambers apparently spiro- loculine, with sharply keeled periphery, later chambers uniserial, triangular in cross-section, edges sharply keeled, widest at the base of each chamber and slightly truncate. Aperture rounded, at the end of a produced and slightly flaring neck. Holotype from sample 45. Dimensions of holotype : Length 0-38 mm., breadth 0-05 mm. Only two specimens were found, but the unusual and distinctive characteristics appear to justify describing it as a new species. The tricarinate form of the later chambers distin- guishes it from other species of the genus. (113) Articulina sagra d’Orbigny. (Plate III, fig. 7.) Articulina sagra d’Orbigny, 1839, in de la Sagra, Hist. pbys. pol. nat. Cuba, “ Foraminiferes ”, p. 183, pi. 9, figs. 23-26 ; Custman, 1944, p. 11, pi. 2, figs. 6-10. Cushman (loc. cit.) considers that records of this species from localities outside the West Indian region are erroneous, but the present specimens compare closely with specimens of A. sagra from La Chorrera, Cuba, and are considered to be conspecific. Common in some of the dredgings of group 3, a few specimens in the deeper- water sample 45. Most of the larger specimens, including the one illustrated, lack the initial series of chambers, which is trilo- culine, costate and usually (but not invariably) smaller in relation to the later series than in Cuban specimens. (114) Articulina queenslandica n. sp. (Plate III, figs. 8, 9 and 10.) Test elongate, slender, commencing with a compressed spiroloculine series of from 1 to 5 chambers after the proloculus, periphery rounded, chambers sometimes taking up a complete whorl. Later chambers, up to 4 in number, are cylindrical, slightly widened at the base and ornamented with low costae, increasing from 4 or 5 on the first uniserial chamber to 12 in the last. Costae increase in height toward the oral end of each chamber and relatively so in successive chambers. Aperture terminal, round, with a narrow everted lip. Holotype from sample 45. Dimensions of holotype : Length 0-42 mm., breadth 0-06 mm. Three specimens were found in'sample 45. The best-developed specimen (fig. 10) which is 0-86 mm. long, would have been referred to N bdobacularia had it been found alone, as the early series consists of a proloculus and one spiral chamber taking up a complete whorl. Another specimen (fig. 9) shows 2 spiroloculine chambers followed by one which takes up nearly one whorl, and a third (fig. 8) which has been selected as the type, has 5 spiroloculine chambers after the proloculus, which is much smaller than in the other two specimens. This species resembles A. pauciloculata Cushman, described from the West Atlantic, in its elongate form and gradation in number of costae, but the latter species has a triloculine initial series of chambers. It is noteworthy that A. queenslandica and A. tricarinata ( vide FORAMIXIFERA— COLLINS 367 supra ) appear to have been derived from a spiroloculine ancestor, rather than triloculine or quinqueloculine as with other species of the genus. However, as Articulina is evidently a polyphyletic genus, it does not appear necessary to make any distinction other than specific. Genus Tubinella Rhumbler, 1906. (115) Tubinella funalis (Brady). Articulina funalis Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 186, pi. xiii, figs. 6-11. Tubinella f unalis (Brady). Rhumbler, 1906, p. 26, pi. 2, fig. 3 ; Cushman, 1924, p. 54, pi. 19, figs. 7, 8. Not uncommon in the samples of group 3. (116) Tubinella inornata (Brady). Articulina funalis Brady, var. inornata Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 186, pi. xiii, figs. 3-5. Tubinella inornata (Brady). Rhumbler, 1906, p. 27, pi. 2, fig. 4 ; Cushman, 1924, p. 54. One specimen of this smooth form was found in sample 45. The globose proloculus is followed by a second spiral chamber, attached and forming half a whorl, from which develops the tubular rectilinear portion. Genus Nubeculina Cushman, 1924. (117) Nubeculina divaricata advena Cushman. Nubeculina divaricata (Brady), var. advena Cushman, 1924, Publ. 342, Carnegie Inst. Washington, p. 53, pi. 19, figs. 1—4. Common in the samples of group 3. All the specimens were referable to this subspecies, which differs from N. divaricata divaricata by the close apposition of the chambers and the clavate form of the test. Genus Hauerina d’Orbigny, 1839. (118) Hauerina diver sa Cushman. Hauerina diversa Cushman, 1946, Contr. Cushman Lab. Foram. Research, 22, pt. 1, p. 11, pi. 2, figs. 16-19. This small species was described from the Paumotus. It occurs rarely in the samples of group 3. (119) Hauerina pacifica Cushman, rugosa n. subsp. (Plate III, figs. 11 a, b, c .) This subspecies differs from II. pacifica pacifica in the development of oblique ridges or crenulations on the later chamber walls. There is considerable variation in the degree of crenulation, some specimens being nearly smooth, others strongly ridged with fine longitudinal striae on the last chamber. Rare in the samples of group 3. Holotype from sample 27. Dimensions of holotype : Length 0-60 mm., breadth 0-56 mm., thickness 0-28 mm. vi. 6. 25 368 GREAT BARRIER REEF EXPEDITION (120) Hauerina fragilissima (Brady). Spiroloculina fragilissima Brady 1884, Rep. Voy. Chall., Zool. 9, p. 149, pi. ix, figs. 12-14. Hauerina fragilissima (Brady). Cushman, 1946, p. 10, pi. 2, figs. 1-6, 8 (synonymy). Not uncommon in the samples of group 3. In the recent state the species appears to be confined to the Pacific. (121) Hauerina involuta Cushman. Hauerina ornatissima (part) Brady, 1884 (not of d’Orbigny), p. 192, pi. vii, figs. 15-17. H. involuta Cushman, 1946, Contrib. Cushman Lab. Foram. Research, 22, pt. 1, p. 13, pi. 2, figs. 25-28. This is the commonest species of the genus in the present material, occurring generally in the samples of groups 1, 2 and 3. It appears to be a restricted Pacific species. (122) Hauerina bradyi Cushman. Hauerina compressa Brady, 1884 (not of d’Orbigny), p. 190, pi. xi, figs. 12, 13 ; Millett, 1898 etc. (1898), p. 610, pi. 13, fig. 11. Hauerina bradyi Cushman 1910 etc. (1917), U.S. Nat. Mus. Bull. 71, pt. 6, p. 62. Frequent in the samples of group 3. It has a general Indo-Pacific distribution. Genus Schlumbergerina Munier-Chalmas, 1882. (123) Schlumbergerina alveoliniformis (Brady). Miliolina alveoliniformis Brady, 1879, Quart. Journ. Micr. Sci. XIX, n.s., p. 268; 1884, p. 181, pi. viii, figs. 15-20. Schlumbergerina alveoliniformis (Brady). Cushman, 1918 etc. (1929), p. 56. Not uncommon in the samples of group 3. Fine specimens were found in sample 26. Genus Ammomassilina Cushman, 1933. (124) Ammomassilina alveoliniformis (Millett). Massilina alveoliniformis Millett, 1898, Journ. Roy. Micr. Soc., p. 609, pi. 13, figs. 5-7 ; Heron-Alien and Earland, 1914 etc. (1915), p. 584, pi. xlv, fig. 15. Ammomassilina alveoliniformis (Millett). Cushman, 1933, p. 32, pi. 3, figs. 5 a, b. Common in samples of group 3, especially those from off-shore stations. Genus Triloculina d’Orbigny, 1826. (125) Triloculina bertheliniana (Brady). Miliolina bertheliniana Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 166, pi. cxiv, figs. 2a, b. M. tricarinata d’Orbigny, reticulated form, Millett, 1898 etc. (1898), p. 503, pi. xi, figs. 12a, b. M. tricarinata d’Orbigny, var. bertheliniana Brady. Chapman, 1900a, p. 174. Triloculina bertheliniana (Brady). Cushman, 1918a, p. 290 ; 1921, p. 457 ; 1932, p. 60, pi. 13, fig. 5. This well-known Indo-Pacific species is common in most of the samples of groups 1, 2 and 3. The larger specimens are not quite so regular as Brady’s figure, tending to develop sinuate edges. FORAinXIFERA— COLLINS 369 (126) Triloculina bicarinata d’Orbigny. Triloculina bicarinata d'Orbigny, 1839, in de la Sagra, Hist. phys. pol. nat. Cuba, “ Foraminiferes ”, p. 158, pi. 10, figs. 18-20 ; Custman, 1922, p. 76, pi. 12, fig. 7 ; 1932, p. 60, pi. 13, figs. 6 a, c. This irregularly-reticulate form with truncate chambers is not uncommon in the samples of groups 1, 2 and 3. Cushman has recorded it from various stations in the tropical Pacific. (127) Triloculina littoralis n. sp. (Plate III, figs. 12a, b, c.) Test porcellaneous, white and polished, ovate in side view with 3 chambers visible, subtriangidar in end view with broadly-rounded periphery. Chambers strongly embracing at ends, costate. Apertural end produced as a short neck, aperture oval with everted lip and large thin tooth. Holotype from sample 7. Dimensions of holotype : Length 0-40 mm., breadth 0-22 mm., thickness 0-16 mm. This small species is common in most of the samples of groups 1, 2 and 3, and is more frequent than usual in the mangrove-pool samples. It resembles the form figured by Heron- Alien and Earland (1915, pi. xlii, figs. 7-9) as Miliolina cultrata Brady, striate form, but differs in its rounded periphery. (128) Triloculina oblonga (Montagu). V ermiculum oblongutn Montagu, 1803, Test. Brit., p. 522, pi. 11, fig. 9. Triloculina oblonga (Montagu). d’Orbigny, 1826, p. 301, no. 16. Elongate smooth triloculine specimens with rounded periphery, large rounded aperture and narrow simple tooth are referred to this species. They are common in the samples of groups 1, 2 and 3. (129) Triloculina quadrata n. sp. (Plate III, figs. 13a, b, c .) Test porcellaneous, 3 chambers visible externally, surface irregularly pitted but shining, ovate in side view, chambers broadly truncate with a rounded keel on each angle, apertural end produced in a short neck, aperture oval with a small simple tooth. Holotype from sample 25. Dimensions of holotype : Length 0-90 mm., breadth 0-65 mm., thickness 0-53 mm. This species is somewhat similar to T. quadrilateralis d’Orbigny, but differs in its produced neck and circular aperture. It also resembles Massilina annectens d’Orbigny, from which it differs in being clearly triloculine, and also in the apertural features. (130) Triloculina sublineata (Brady) ? (Plate III, figs. 14a, b.) ? Miliolina circularis Bornemann, var. sublineata Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 169, pi. iv, figs. 7a-c. M. circularis Bornemann, var. sublineata Brady. Millett, 1898 etc. (1898), p. 501, pi. xi, fig. 4 ; Heron- Alien and Earland, 1914 etc. (1915), p. 558, pi. xli, figs. 9-11. All the specimens found were thin-shelled and translucent, with the cribrate aperture figured by Millett. Whether this is the same form as Brady’s var. sublineata is open to doubt. 370 GREAT BARRIER REEF EXPEDITION Brady’s figure shows an arcuate aperture without a tooth, and appearing to he closed by what might be interpreted as a finely cribrate plate (fig. 7c), but his description does not mention this. Heron- Allen and Earland’s fig. 1 1 shows an oval cribrate aperture which may be somewhat conventionalized. The question can only be resolved by the examination of Brady’s figured specimens, and is therefore left open. The species is not uncommon in the samples of groups 1, 2 and 3. (131) Triloculina subrotunda (Montagu). Vermiculum subrotundum Montagu, 1803, Test. Brit., p. 521. Triloculina subrotunda Parr, 1950, p. 293 (synonymy). Typical examples were fairly common in the shallow- water samples of groups 1 and 2. (132) Triloculina terquemiana (Brady). Miliolina terquemiana Brady, 1884, Rep. Yoy. Chall., Zool. 9, p. 166, pi. cxiv, figs, la, b. The degree of striation in this species appears to be variable, also the degree of inflation of the chambers, as noted by Heron- Allen and Earland (1915, p. 563). It is common through- out the samples of groups 1, 2 and 3. Records range from the western Indian Ocean to the Ellice Islands. (133) Triloculina transversistriata (Brady). Miliolina transversistriata Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, n.s., p. 45 ; 1884, Rep. Yoy. Chall., Zool. 9, p. 177, pi. iv, figs. 6a. c. This well-marked species occurred in small numbers in many of the samples of groups 1 and 2. (134) Triloculina tricarinata d’Orbigny. Triloculina tricarinata d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 299, no. 7, Modeles no. 94. Specimens referable to this wide-ranging species were frequent in most of the samples of groups 1, 2 and 3. Genus Edentostomina, n. gen. Test triloculine or rarely biloculine, compressed, ovate in outline, last 2 chambers added at approximately 180°, periphery sharp or carinate, aperture oval with a thickened rim and no tooth. Genotype Miliolina cultrata Brady. Millett (1898), in describing Miliolina durrandii, states “ This is one of an interesting group in which the aperture is a large elliptical or fusiform opening without teeth ”, and includes M. cultrata and M. ruperticma in this edentate group. Heron- Allen and Earland, 1914 etc. (1915) group the same species together as the “ Group of M. cultrata (edentate) ”. This appears to be a useful natural grouping which should be recognized in nomenclature, the main point of difference from Triloculina s. str. being the large oval and edentate aperture. It is considered that Pyrgo milletti Cushman, which was originally described as a variety of Miliolina durrandii, should also be included in this genus, as its affinities appear to be closer to this group than to any described species of Pyrgo. The species included in this genus appear to be confined to the Indo-Pacrfic region, where the genus has developed in Recent times. FORAMIXIFERA — COLLINS 371 (135) Edentostomina cultrata (Brady). Miliolina cultrata Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, n.s., p. 45 ; 1884, p. 161, pi. v, figs. 1, 2. Found in small numbers in many of the samples of group 3. (136) Edentostomina durrandii (Millett). Miliolina durrantii Millett, 1898, Journ. Roy. Micr. Soc., p. 268, pi. vi, figs. 7-16 ; Heron-Alleu and Earland, 1914 etc. (1915), p. 565, pi. xlii, figs. 11-16. Not uncommon in the samples of group 3. Practically all the specimens are striate, only the immature tests being smooth, and then having slight striations near the apertural end of the chamber. (137) Edentostomina rupertiana (Brady). Miliolina rupertiana Brady 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, n.s., p. 46 ; 1884, p. 178, pi. vii, figs. 7-12 ; Millett, 1898 etc. (1898), p. 269, pi. vi, fig. 13 ; Heron-Alien and Earland, 1914 etc. etc. (1915), p. 565. This beautifully decorated species was fairly common in the samples of group 3, some very large specimens being found. (138) Edentostomina milletti (Cushman). Miliolina durrandii Millett (part), 1898, Journ. Roy. Micr. Soc., p. 268, pi. vi, figs. 8-10 (not 7). Biloculina milletti Cushman, 1910 etc. (1917), XJ.S. Nat. Mus. Bull. 71, 4, p. 479. Pyrgo milletti (Cushman). Cushman, 1918 etc. (1929), p. 68, pi. 19, fig. 1. Biloculine specimens referable to Cushman’s species occurred in the samples of group 3. In their compressed form, keeled edge and oval toothless aperture they resemble other species in this group, in which some specimens occur with the antepenultimate chamber only just visible between the embracing later chambers. The small increase in overlap of the last chamber needed to make the test biloculine instead of triloculine is considered to be of little weight as a character compared with the points of resemblance with other species of the genus, and it is therefore included in Edentostomina. Genus Pyrgo Defrance, 1824. (139) Pyrgo striolata (Brady). Biloculina ringens (Lamarck), var. striolata Brady, 1884, Rep. Yoy. Chalk, Zool. 9, p. 143, pi. iii, figs. 7, 8. Occasional specimens were found in samples of all groups, but it was best represented in the deeper-water samples of group 4. (140) Pyrgo denticulata (Brady). Biloculina ringens (Lamarck), var. denticulata Brady, 1884, Rep. Voy. Chalk, Zook 9, p. 143, pi. iii, figs. 4, 5. Pyrgo denticulata (Brady). Cushman, 1918 etc. (1929), p. 69, pi. xviii, figs. 3, 4 ; Parr, 1950, p. 296. Occasional specimens were found in the samples of groups 1 , 2 and 3, and it was common in sample 53. Specimens without the basal serration occur, as noted by Parr in southern 372 GREAT BARRIER REEF EXPEDITION Australian waters, but these are otherwise inseparable from serrate specimens in the same material. (141) Pyrgo subglobulus Parr. Pyrgo subglobulus Parr, 1950, B.A.N.Z.A.R. Exped. Rept. Y, pt. 6, Foram., p. 298, pi. vii, figs. 10a. c. Four specimens were found in sample 45. This species was described from off Tasmania. (142) Pyrgo depressa (d’Orbigny). Biloculina depressa d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 298, no. 7 ; Modeles no. 91 ; Brady, 1884, p. 145, pi. ii, figs. 12, 15-17, pi. iii, figs. 1, 2. Pyrgo depressa (d’Orbigny). Cushman, 1918 etc. (1929), pt. 6, p. 71, pi. 19, figs. 4, 5. This wide-ranging species occurred fairly frequently in the samples of groups 1, 2 and 3. Family Ophthalminidae. Genus Cornuspira Schultze, 1854. (143) Cornuspira involvens Reuss. Operculina involvens Reuss, 1850, Denkschr. d. k. Akad. Wiss. Wien, 1, p. 370, pi. xiv, fig. 20. Cornuspira involvens Reuss, 1863, Sitzungsb. d. k. Akad. Wiss. Wien, XLVIII, p. 39, pi. i, fig. 2 ; Brady, 1884, p. 200, pi. xi, figs. 1-3. Frequent in the finer fractions of the samples of groups 1, 2 and 3. Specimens were invariably small, not more than 0-25 mm. in diameter, with 4 to 5 convolutions. Genus Cornuspirella Cushman, 1928. (144) Cornuspirella diffusa (Heron-Alien and Earland). (Plate III, fig. 16.) Cornuspira diffusa Heron-Alien and Earland, 1913, Journ. Roy. Micr. Soc., pp. 272-276, pi. 12 ; Earland, 1933, p. 51, pi. 1, figs. 5-7 ; 1934, p. 52. Cornuspirella diffusa (Heron- Allen and Earland). Cushman, 1928a, p. 4, pi. 1, fig. 14. Two fragments of tests, from samples 24 and 25 respectively, are referable to this widely-distributed species, which was described from the eastern North Atlantic and is recorded also from the Antarctic. The occurrence is noteworthy, since records are generally from cold-water areas. Genus N odophthalmidium MacFadyen, 1939. (145) N odophthalmidium simplex Cushman and Todd. N odophthalmidium simplex Cushman and Todd, 1944, Contrib. Cushman Lab. Foram. Research, 20, pt. 3, p. 67, pi. 11, fig. 8. Five specimens from sample 45, lacking initial chambers, are referable to this species. One has 4 elongate-pyriform chambers and reaches a length of 1 ' 1 mm. (146) N odophthalmidium gracilis n. sp. (Plate III, fig. 15.) Articulina conico-articulata Millett 1898 (not of Batsch), Journ. Roy. Micr. Soc., p. 511, pi. xii, figs. 9, 10a-c. Test elongate, slightly curved, circular in cross-section, early portion consisting of a spherical proloculus and a short planispiral tube of less than one whorl in length, merging FORAMINIFERA — COLLIXS 373 into a rectilinear series of long slender chambers, widest just above the suture and tapering to the oral end, wall ornamented by 8-10 high sharp costae, in some specimens terminated by a basal spine, aperture terminal, round, with an everted lip. Holotype from sample 45. Dimensions of holotype : Length 0-75 mm., breadth 0-12 mm. This is the species figured by Millett and later referred by Cushman to his N. milletti, described from Fiji. It differs from the latter species in the consistently longer and more slender chambers with fewer costae. Common in sample 45 and occasional in the samples of group 3. Genus N odobacu lariella Cushman and Todd. (147) Nodobaculariella rustica Cushman and Todd. NodobacularieRa rustica Cushman and Todd, 1944, Contrib. Cushman Lab. Foram. Research, 20, pt. 3, p. 73, pi. 12, figs. 4, 5. This species, described from the Phillippines and recorded also from the Paumotus, was rare in sample 26. Genus Vertebralina d’Orbigny, 1826. (148) Vertebralina striata d’Orbigny. Vertebralina striata d’Orbigny, 1826, Ann. Sci. Nat. XII, p. 283 ; Modeles no. 81. This widely-distributed species was not uncommon in the samples of group 3. Genus Ophthalmidium Zwingli and Kubler, 1870. (149) Ophthalmidium inconstans (Brady). Hauerina inconstans Brady, 1879 etc., Quart, Journ. Micr. Sci. XIX, n.s., p. 54. Ophthalmidium inconstans (Brady). Brady, 1884, p. 189, pi. xii, figs. 5, 7, 8. One large winged specimen was found in sample 45. (150) Ophthalmidium circularis (Chapman), tropicalis n. subsp. (Plate IV, figs, la, b.) Test subcircular, commencing with a proloculus and a short spiral tube, then becoming septate and irregularly spiroloculine. Chambers narrow and deep, connected to the chambers of earlier whorls by a thin plate which is discontinuous, being broken by gaps along the line of attachment to earlier whorls. Periphery slightly concave, chambers irregular in outline, apertural end tapering to a small round aperture at the end of a long neck. Holotype from sample 45. Dimensions of holotype : Diameter 0-25 mm., thickness 0-10 mm. This subspecies differs from O. circularis circularis (Chapman, 1915, p. 7, pi. 1, fig. 1 ; 1941, p. 186) in the more angular and irregular outline of the later chambers and in the discontinuity of the plate connecting successive whorls. Young specimens are hardly separable from Chapman’s species, but the adult test is quite distinctive. Eleven specimens were found, all in sample 45. 374 GREAT BARRIER REEF EXPEDITION Genus Polysegmentina Cushman, 1946. (151) Polysegmentina circinata (Brady). Hauerina circinata Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, n.s., p. 17 ; 1884, p. 191, pi. xi, figs. 14-16. Polysegmentina circinata (Brady). Cushman, 1946, p. 1, pi. 1, figs. 1-4 (synonymy). This species was common in the shallow-water samples, particularly in those of group 3. It is a characteristic species of the muddy coastal waters of North Queensland. Genus Planispirinella Wiesner, 1931. (152) Planispirinella exigua (Brady). Planispirina exigua Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 196, pi. xii, figs. 1-4, text-figs. 5, 6. Planispirinella exigua (Brady). Wiesner, 1931, p. 69. Common in the samples of groups 3 and 4. (153) Planispirinella involuta n. sp. (Plate IV, figs. 2 a, b.) Test porcellaneous, planispiral, with 3 chambers making up the last whorl, compressed and discoidal, thickest at the centre, involute. Sutures radial, straight, recurved near periphery, very slightly depressed. Chambers commence with a proloculus followed by a spiral tube of 2-3 whorls, then becoming septate with 3 chambers to the whorl. The chambers are tubular and peripheral, but have alar extensions reaching to the centre of the test and covering the earlier whorls ; the hollow tubular portion of the last 3 chambers appears as a semi-translucent border. The aperture is oblique to the periphery and consists of the open end of the tubular chamber. Holotype from sample 45. Dimensions of holotype : Diameter 0-41 mm., thickness 0-09 mm. This species is distinct from P. exigua (Brady) which occurs in the same material, the difference lying in the more circular outline and the involute character of the later chambers. Four specimens were found in sample 45 only. Genus Wiesnerella Cushman, 1933. (154) Wiesnerella auriculata (Egger). Planispirina auriculata Egger, 1893, Abhandl. kon. bay. Akad. Wiss. Muncben, cl. ii, 18, Abt. 2, p. 245, pi. iii, figs. 13, 15 ; Heron-Alien and Earland, 1914 etc. (1915), p. 590, pi. xlvi, figs. 3-7. Ophthalmidium cornu Chapman, 1901, p. 408, pi. 36, fig. 1. Wiesnerella auriculata (Egger). Cushman, 1933, p. 33. Not uncommon in the samples of groups 1, 2 and 3. Chapman’s species from Funafuti is apparently the same as Egger’s earlier-described form. A closely allied species is found in the Miocene clays of Batesford, Victoria. FORAMIXIPERA— COLLINS 375 Genus Nubecularia Defrance, 1825. (155) Nubecularia decorata Heron- Allen and Earland. Nubecularia lucifuga Defrance, var. decorata Heron- Allen and Earland, 1914 etc. (1915), Trans. Zool. Soc. London, 20, pt. 17, p. 549, pi. xl, figs. 6, 7. A small form occurring attached to shell fragments in sample 2 is referred to the above species, though the surface appears to be granular rather than pustulate. The invariably rectilinear growth of this species is sufficient to separate it from N . lucif uga, which occurs commonly in the coastal waters of southern Australia. Genus Nubeculopsis n. gen. Test porcellaneous, adherent, chambers low-domed, commencing with a proloculus followed by 2 reniform chambers each taking up nearly half a whorl, continuing with a a series of arcuate chambers in a spiral which may reverse its direction after the first 3 or 4 chambers, becoming irregularly uniserial or with tubular chambers piled above the earlier spiral series. Aperture in the early chambers a simple opening at the distal end of the chambers, in the later tubular chambers a flaring lipped opening of the full width of the chamber. This genus differs from Nubecularia in being septate throughout, without any cornu- spirine spiral tube. (156) N iibeculopsis queenslandica n. sp. (Plate IV, figs. 3a, b, c .) Species with the characters of the genus. Holotype from sample 40, at which the species is rare. Dimensions of holotype : Greatest diameter 0-82 mm. Genus Panina Cushman, 1931. (157) Panina bradyi (Millett). Nubecularia injlata Brady 1884 (not of Terquem), p. 135, pi. i, figs. 5-8. N. bradyi Millett, 1898, Journ. Roy. Micr. Soc., p. 261, pi. v, figs. 6 a, b. Parrina bradyi (Millett). Cushman, 1931, p. 20. Not uncommon in the samples of groups 1, 2 and 3. Family Peneroplidae. Genus Peneroplis Montfort, 1808. (158) Peneroplis planatus (Fichtel and Moll). Nautilus jplanatus, var. /?, Fichtel and Moll, 1798, Test. Micr., p. 91, pi. xvi, figs. 1 d, e,f. Peneroplis planatus (Fichtel and Moll). d’Orbigny, 1826, p. 285, no. 1 ; Modeles no. 16. Common in the samples of groups 1, 2 and 3. This species is widely distributed on the Australian coast, except for the cooler waters of Bass Strait. 376 GREAT BARRIER REEF EXPEDITION (159) Peneroplis pertusus (Forskal). Nautilus pertusus Forskal, 1775, Descr. Anim., p. 125, no. 65. Peneroplis pertusus (Forskal). Jones, Parker and Brady, 1865, p. 19. Present in most of the samples of groups 1, 2 and 3, hut not so common as P. planatus. (160) Peneroplis arietinus (Batsch). Nautilus ( Lituus ) arietinus Batsch, 1791, Conch. Seesandes, p. 4, pi. vi, fig. 15c. Peneroplis arietinus (Batsch). Parker, Jones and Brady, 1865, p. 26, pi. 1, fig. 18. Well-grown specimens were fairly common in the samples of group 3. Genus Spirolina Lamarck, 1804. (161) Spirolina cylindraceus Lamarck. Spirolina ( Spirolinites ) cylindraceus Lamarck, 1804, Ann. Mus. V, p. 245, No. 2. Fine specimens were found in the samples of group 3, with as many as 18 chambers in the rectilinear portion. Occasional specimens occurred in the samples of groups 1 and 2. Genus Monalysidium Chapman, 1900. (162) Monalysidium politum Chapman. Peneroplis ( Monalysidium ) polita Chapman, 1900, Journ. Linn. Soc., Zool. 28, p. 4, pi. 1, fig. 5. Monalysidium polita Chapman. Heron- Allen and Earland, 1914 etc. (1915), p. 603, text-fig. 43#. Occasional specimens were found in the samples of groups 1, 2 and 3. Specimens with as many as 10 chambers were found, but in each case the earliest chamber was exactly the same as the rest, and perforated at the base. Genus Sorites Ehrenberg, 1839. (163) Sorites marginalis (Lamarck). Orbulites marginalis Lamarck, 1816, Hist Nat. Anim. sans Vert. II, p. 196, No. 1. Sorites marginalis (Lamarck). Cushman, 1918 etc. (1930), p. 49, pi. 18, Figs. 1-4. Bather rare in the samples of group 3. Specimens up to 3- 9 mm. in diameter were found. A few juvenile specimens were present in the deeper- water sample 45. Genus Amphisorus Ehrenberg, 1840. (164) Amphisorus duplex (Carpenter). Orbitolites duplex Carpenter, 1883, Zool. Chall. Exped., pt. XXI, p. 25, pi. iii, figs. 8-14 ; pi. iv, figs. 6-10 ; pi. v, figs. 1-13 ; Brady, 1884, p. 216, pi. xiv, fig. 7. Amphisorus duplex (Carpenter). Said, 1949, p. 25, pi. 3, fig. 1. This species, recently re-established by Said for forms which differ from the genotype (A. hemprichii Ehrenberg) in being thinner, with rounded periphery and oval chambers, is usually present in small numbers together with Marginopora vertebralis (vide infra) in the samples of groups 1, 2 and 3. The record by the author of A. hemprichii from the Pleistocene of western Victoria (Collins, 1953) was erroneous, and should refer to the present FORAMESilFERA — COLLINS 377 species. It has, however, not been recorded from southern Australian waters, and its southward limit in the coastal waters of eastern Australia does not appear to have yet been established. Genus Marginopora Blainville, 1830. (165) Marginopora vertebralis Quoy and Gaimard. Marginopora vertebralis Quoy and Gaimard. 1833, Yoy. de Astrolabe, fide Blainville 1834, Man. d’Actinologie, p. 412, pi. lxix, figs. 6a-c ; Cushman, 1932 etc. (1933), p. 67, pi. 19, figs. 11, 12. Orbitolites complanata of authors (not of Lamarck). This is the commonest large foraminifer of the Barrier Reef islands, forming a major constituent of the beach sands of the region. So well known is it that the tests are dyed in various colours and strung into necklaces for sale to tourists visiting the island resorts. It is recorded from the Upper Pliocene of Adelaide and the Pleistocene of York Penin- sula, South Australia. Chapman and Parr have recorded it in the Recent state from the Great Australian Bight, but it does not occur in the cooler waters of the south-east coast. Its southward limits on the east coast do not appear to have been determined. Specimens in the present material range up to 2-2 cm. in diameter (sample 31) but the usual shore- sand specimens rarely reach more than 1 cm. Family Alveolinidae. Genus Alveolinella H. Douville, 1906. (166) Alveolinella quoyi (d’Orbigny). Alveolina quoyi d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 306, pi. 17, figs. 11, 13. Alveolinella quoyi (d’Orbigny). Douville, 1906, p. 585. Alveolina boscii of authors (not of Defrance). Both the megalospheric and microspheric forms were found in the samples of group 3, the former being much more common. It was not usually present in the beach sands, possibly because its elongate shape resists transportation by waves and currents to a greater degree than the discoidal forms which are predominant in the coarser fractions of these samples. Occasional juvenile specimens were found in the deeper-water samples of group 4, where local currents had probably transported them in the embryonic stage. Family Fischerinidae. Genus Fischer ina Terquem, 1878. (167) Fischerina pellucida Millett. (Plate IV, fig. 4.) Fischerina pellucida Millett, 1898, Journ. Roy. Micr. Soc., p. 611, pi. xiii, figs. 14, 15. Two specimens were found in sample 45. One has the shell wall finely pitted and may be distinct. 378 GREAT BARRIER REEF EXPEDITION Superfamily LAGENIDEA. Family Lagenidae. Genus Lagena Walker and Jacob, 1798. (168) Lagena ampulla-distoma Rymer Jones. Lagena vulgaris Williamson, var. ampulla-distoma Rymer Jones, 1872, Trans. Linn. Soc., London, XXX, p. 63, pi. xix, fig. 52. Rare, one specimen from sample 44. (169) Lagena auriculata Brady ? ? Lagena auriculata Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, n.s., p. 61 ; 1884, p. 487, pi. lx, figs. 29, 31, 33. L. auriculata Brady. Sidebottom, 1912, p. 420, pi. 20, figs. 13, 14. One specimen of the variety illustrated by Sidebottom in figs. 13 and 14 was found. This variety has very little in common with Brady’s fig. 29, which, as the first illustrated, must presumably be taken as the type of L. auriculata. It appears to be closer to L.formosa Schwager, from which it differs only in the splitting of the single keel to form a loop-shaped double keel. (170) Lagena chasten Millett. Lagena chasten Millett, 1901, Journ. Roy. Micr. Soc., p. 11, pi. 1, fig. 11. Typical specimens were found in sample 23. (171) Lagena elongata (Ehrenberg). Miliola elongata Ebrenberg, 1854, Mikrogeologie, pi. xxv, fig. 1. Lagena elongata (Ebrenberg). Brady, 1884, p. 457, pi. lvi, fig. 29. This elongate, parallel-sided species was frequent in the finer fractions of the shallow- water samples of groups 1, 2 and 3. (172) L. gracillima (Seguenza). Amphorina gracillima Seguenza, 1862, Foram. Monotal. Messina, p. 51, pi. 1, fig. 31. Lagena gracillima (Seguenza). Jones, Parker and Brady, 1866, p. 45, pi. 1, figs. 36, 37. Fusiform specimens, widest at the middle and tapering conically toward the extremities and having a phialine neck and lip, are referred to this species. They were common in the samples of groups 1, 2 and 3. (173) Lagena laevis (Montagu). Vermiculum laeve Montagu, 1803, Test. Brit., p. 524. Lagena laevis (Montagu). Williamson, 1848, p. 12, pi. 1, figs. 1, 2. Smooth bottle-shaped specimens with rounded base and phialine neck and lip are referred to this species. They were occasional in the samples of group 3. FORAMINIFERA — COLLINS 379 (174) Lagena pacifica Sidebottom. Lagena pacifica Sidebottom, 1912, Journ. Quekett ISlicr. Club, II, p. 398, pi. 16, fig. 29. One specimen identical with Sidebottom’s figure was found in sample 45. (175) Lagena perlucida (Montagu). Y ermiculum perlucidum Montagu, 1803, Test. Brit., p. 525, pi. 14, fig. 3. Lagena perlucida (Montagu). Cushman and McCulloch, 1950, p. 342, pi. 46, figs. 1, 2 (synonymy). Common in samples of groups 1, 2 and 3. (176) Lagena spiralis Brady. Lagena spiralis Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 468, pi. cxiv, fig. 9. L. striato- punctata Parker and Jones, var. spiralis Brady. Millett, 1898 etc. (1901), p. 489, pi. viii, fig. 7 ; Cushman and McCulloch, 1950, p. 353, pi. 47, figs. 17, 18 (synonymy). Occasional specimens were found in the samples of group 3. Quite similar specimens are found in the Miocene of Victoria, and the consistent variation of this form from L. striato- punctata s. str. over a long period of thne appears to justify its specific rank as originally described by Brady. (177) Lagena striato-punctata Parker and Jones. Lagena striata (d'Orbigny), var. striato-punctata Parker and Jones, 1865, Phil. Trans. CLV, p. 350, pi. xiii, figs. 25-27. L. striato-punctata Parker and Jones. Brady, 1884, p. 468, pi. lviii, figs. 37, 40. Rare in sample 45 (one specimen). (178) Lagena striata (d’Orbigny). Oolina striata d’Orbigny, 1839, Foram. Amer. Merid., p. 21, pi. v, fig. 12. Lagena striata (d’Orbigny). Brady, 1884, p. 460, pi. lvii, figs. 22, 24, 28, 29 etc. Flask-shaped, finely- striate specimens with an apieulate base and a produced annulate neck have been referred to this species. This is perhaps the commonest species of the genus in the samples of groups 1, 2 and 3. (179) Lagena sulcata Walker and Jacob. Serpula ( Lagena ) sulcata Walker and Jacob, 1798, Adams’ Essays, Kanmacher’s Edn., p. 634, pi. xiv, fig. 5. Flask-shaped specimens with bold costae set widely apart have been referred to this species. They are common in samples of all groups. Genus Oolina d’Orbigny, 1839. (180) Oolina hexagcma (Williamson). Entosolenia squamosa (Montague) var. hexagona Williamson, 1848, Ann. Mag. Nat. Hist., ser. 2, p. 20, pi. ii, fig. 23. Oolina hexagona (Williamson). Parr, 1950, p. 304. Rare in sample 45. 380 GREAT BARRIER REEF EXPEDITION (181) Oolina pseudocatenulata (Chapman and Parr). Lagena pseudocatenulata Chapman and Parr, 1937, Aust. Antarctic Exped. 1911-14, 1, pt. 2, 332, pi. vi, figs. 75, 76. Oolina pseudocatenulata (Chapman and Parr). Parr, 1950, p. 301, pi. viii, fig. 5. Common in sample 45. This species was described from the Antarctic. (182) Oolina squamosa (Montagu). V ermiculum squamosum Montagu, 1803, Test. Brit., p. 526, pi. xiv, fig. 2. Oolina squamosa (Montagu). Parr, 1950, p. 303. Rare in sample 45. Genus Fissurina Reuss, 1850. (183) Fissurina contusa Parr. Fissurina contusa Parr, 1945, Proc. Roy. Soc. Victoria, 56, n.s., pt. ii, p. 203, pi. xix, fig. 6. This is the form recorded by Brady (1884) from Raine Island, as ? Lagena castrensis Schwager. One specimen from sample 23 is closely similar to Brady’s fig. 3, pi. cx, on which Parr’s species was based. (184) Fissurina clathrata (Brady). Lagena clathrata Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 485, pi. lx, fig. 4. Fissurina clathrata (Brady). Parr, 1950, p. 310. Two good examples from sample 45. (185) Fissurina formosa brevis (Brady). Lagena for mosa Schwager, var. brevis Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 484, pi. cxiv, fig. 10. One specimen of this beautiful form was found in sample 45. It was described from Raine Island. It is considered here as a subspecies of F. formosa, which in the typical form does not occur in this material, but may prove to be a separate species. (186) Fissurina lacunata (Burrows and Holland). Lagena castrensis Brady (not of Schwager), 1884, Rep. Voy. Chall., Zool. 9, p. 485, pi. lx, figs. 1, 2. L. lacunata Burrows and Holland, 1895, in Jones, Pal. Soc., vol. for 1895, p. 205, pi. 7, figs. 12a, b. Fissurina lacunata (Burrows and Holland). Parr, 1945, p. 203. Specimens similar to Brady’s fig. 2, which was drawn from a Raine Island specimen, occur occasionally in the samples of groups 3 and 4. The species ranges as far south as Bass Strait. In sample 45 there were several examples of the form in which the surface pitting is replaced by a delicate hexagonal network merging into costae, vide Cushman 1932 etc. (1933), p. 27, pi. 7, figs. 2, 3, 4 and 5. Insufficient material was available to confirm Cushman’s opinion that this form intergrades with the typical pitted form. (187) Fissurina lagenoides (Williamson). Enlosolenia marginata Walker and Boys, var. lagenoides Williamson, 1858, Rec. Foram. Gt. Brit., p. 11, pi. i, figs. 25, 26. Lagena lagenoides (Williamson). Brady, 1884, p. 479, pi. lx, figs. 6, 7, 9 and 12. Fissurina lagenoides (Williamson). Parr, 1947, p. 122. Rare in sample 29 (1 specimen). FORAMDIIFERA — COLLINS 381 (188) Fissurina sublagenoides (Cushman). Lagena lagenoides (Williamson). Brady, 1884, Rep. Yoy. Chall., Zool. 9. (pars), pi. lx, figs. 13, 14. L. sublagenoides Cushman, 1910 etc. (1913), U.S. Nat. Mus. Bull. 71, pt. 3, p. 40, pi. 16, fig. 4. Occasional specimens were found in samples of groups 3 and 4. (189) Fissurina lucida (Williamson). Entosolenia marginata Walker and Boys var. lucida Williamson, 1858, Rec. Foram. Gt. Brit., p. 10, pi. 1, figs. 22, 23. Lagena lucida (Williamson). Sidebottom, 1904 etc. (1906), p. 6, pi. i, figs. 9.12. Common in the finer fractions of samples of groups 1, 2 and 3. Specimens are of the type illustrated by Sidebottom’s figs. 10a, b. (190) Fissurina marginato-perforata (Seguenza). Lagena marginato- perforata Seguenza, 1880, Atti. Accad. Lincei, ser. 3, 6, p. 332, pi. 17, fig. 34 ; Cushman, 1932 etc. (1933), p. 19, pi. 5, figs, la, b, pi. 6, figs, la, b , pi. 4, figs. 13a, b. Frequent in samples of all groups. (191) Fissurina orbignyana Seguenza. Fissurina orbignyana Seguenza, 1862, Foram. Monotal. Messina, p. 66, pi. ii, figs. 25, 26. Xot uncommon and generally distributed in all groups. (192) Fissurina walleriana (Wright). Lagena orbignyana (Seguenza) var. walleriana Wright, 1886, Proc. Roy. Irish Acad., ser. 2, 4, p. 611. Rare in sample 45 (2 specimens). (193) Fissurina radiato-marginata (Parker and Jones). Lagena radiato-marginata Parker and Jones, 1865, Phil. Trans. CLV, p. 355, pi. xviii, figs. 3a, b. One example of this species was found in sample 45. (194) Fissurina semistriata (Uchio). Entosolenia semistriata Uchio, 1950, Assoc. Petr. Technol. Journ. (Tokyo). A compressed, pyriform, keeled Fissurina with 6 or more fine costae curving up from the base for about one-third the height of the test is referred to this species, which was described from the Upper Pliocene of Japan. The generic name Entosolenia was shown by Parr (1945, p. 119) to be a synonym of d’Orbigny’s Oolina. Rare in sample 45. (195) Fissurina staphyllearia Schwager. Fissurina staphyllearia Schwager, 1866, Novara-Exped., Geol. Theil. 2, p. 209, pi. 5, fig. 24. Three typical specimens from sample 45. 382 GREAT BARRIER REEF EXPEDITION (196) Fissurina wrightiana (Brady). Lagena wrightiana Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, n.s., p. 62 ; 1884, p. 482, pi. lxi, figs. 6, 7. Rare in sample 32. (197) Fissurina varioperforata (Buchner). Lagena varioperforata Buchner, 1940, Nova Acta Leop. 9, 62, p. 494, pi. 18, figs. 353-358. This species is characterized by the coarse tubulation of the wall, leaving a narrow elliptical area of clear shell- wall in the centre of the test, through which can be seen a straight entosolenian tube extending about half-way down the test. Specimens from sample 45 appear to fall within the limits of variation illustrated by Buchner. The type was from the Mediterranean. Genus Parafissurina Parr, 1945. (198) Parafissurina unguis (Heron-Alien and Earland). Lagena unguis Heron-Alien and Earland, 1913a, Proc. Roy. Irish Acad. 31, p. 86, pi. 7, figs. 1-3. One specimen of this wide-ranging but rare species was found in sample 45. Genus Nodosaria Lamarck, 1812. (199) Nodosaria calomorpha Reuss. Nodosaria calomorpha Reuss, 1865, Denschr. d. k. Akad. Wiss. Wien. XXV, p. 129, pi. 1, figs. 15-19. Eight specimens of this small but distinctive species were found in sample 45. (200) Nodosaria proxima Silvestri. N odosaria proxima Silvestri, 1872, Nodos. Foss, e Yiv. d’ltal., p. 63, pi. vi, figs. 138-147. Two-chambered nodosarians with 8-10 costae occur rarely in the samples of groups 3 and 4, and are referred to this species, which was recorded from Raine Island by Brady. (201) Nodosaria pyrula d’Orbigny. Nodosaria pyrula d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 253, no. 13. Eight specimens were found in sample 45. Several show the elongate fusiform initial chamber which is usually missing from the fragile test of this species. Genus Dentalina d’Orbigny, 1826. (202) Dentalina antarctica Parr. Dentalina antarctica Parr, 1950, B.A.N.Z.A.R. Exped. Rep. V, pt. 6, p. 320, pi. xi, figs. 26, 27. One specimen from sample 45 is very close to the megalospheric specimen illustrated by Parr (fig. 27). (203) Dentalina calif ornica Cushman and Gray. Dentalina californica Cushman and Gray, 1946, Contrib. Cushman Lab. Foram. Res. 22, pt. 2, p. 66. pi. 12, figs. 3-5. A slender form with oblique chambers and a long acuminate proloculus is referred to this species, described from the Pliocene of Timms Point, California. Rare in sample 45. FORAMINIFERA — COLLINS 383 (204) Dentalina inflexa (Reuss). Nodosaria inflexa Reuss, 1865, Denkschr. d. k. Akad. Wiss. Wien, XXV, p. 131, pi. ii, fig. 1 ; Brady, 1884, p. 498, pi. lxii, fig. 9. Two specimens from sample 45 having elongate pyriform chambers but lacking a connecting neck between the chambers appear to be referable to this species. Genus Vaginulina d'Orbigny, 1826. (205) Vaginulina bassensis Parr. Vaginulina bassensis Parr, 1945, Proc. Roy. Soc. Victoria, 56. N.s., pt. 2, p. 200, pi. 12, figs. 4a, b. One specimen was found in sample 45, comparing well with specimens from the type locality on the shores of Bass Strait. Genus Marginulina d'Orbigny, 1826. (206) Marginulina glabra d'Orbigny. Marginulina glabra d’Orbigny", 1826, Ann. Sci. Nat. VII, p. 259, no. 6, Modeles no. 55. Rare in sample 45 (2 specimens). Genus Amphicorgne Schlumberger, 1881. (207) Amphicorgne scalaris (Batscli). Nautilus (Orthoceras) scalaris Batsch, 1791, Conchyl. Seesandes, No. 4, pi. ii, figs. 4a, b. Nodosaria scalaris (Batsch). Brady, 1884, p. 510, pi. lxiii, figs. 28-31. Eight specimens were found in sample 45, all megalospheric. The writer has followed Parr (1950) in referring this species to the genus Amphicorgne. One specimen had the final chamber separated from the earlier ones by the full length of the apertural neck, a condition rather similar to Brady’s variety separans, in which, however, the final chamber commences at the base of the apertural neck of the penultimate chamber. Such forms occur, in the writer’s experience, in any large series of specimens of A. scalaris , and therefore represent an individual modification rather than a geographical variety or subspecies, as suggested by Brady. (208) Amphicorgne hirsuta (d’Orbigny). Nodosaria hirsuta d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 252, no. 7. Amphicorgne hirsuta (d’Orbigny). Parr, 1950, p. 328. Two specimens were found in sample 45, both megalospheric. Genus Planularia Defrance, 1824. (208) Planularia sp. alf. P. tricarinella (Reuss). Cristellaria tricarinella Reuss. Brady, 1884, Rep. Voy. Chalk, Zool. 9, p. 540, pi. lxviii, figs. 3 and 4. One specimen from sample 45. This form has 17 chambers after the proloculus, and its dorsal edge is tricarinate as figured by Brady. It is not referable to P . australis Chapman (1941), which has fewer chambers and a rounded periphery, vi. 6. 26 384 GREAT BARRIER REEF EXPEDITION Genus Lenticulina Lamarck, 1804. (210) Lenticulina iota (Cushman). Cristellaria iota Cushman, 1918 etc. (1923), U.S. Nat. Mus. Bull. 104, pt. 4, p. Ill, pi. xxix, fig. 2, pi. xxx, fig. 1. Lenticulina iota (Cushman). Parr, 1950, p. 320. Specimens are not uncommon in sample 45. As noted by Parr in the case of a Tasmanian specimen, the chambers are rather more recurved than in Cushman’s figures. (211) Lenticulina peregrina (Sch wager). Cristellaria peregrina Schwager, 1866, Novara-Exped., Geol. Theil., p. 245, pi. vii, fig. 89. Lenticulina peregrina (Schwager). Parr, 1950, p. 321. Three good examples were found in sample 45. Subgenus Robulus Montfort, 1808. (212) Lenticulina ( Robulus ) altifrons Parr. Cristellaria acutiauricularis Brady (not of Fichtel and Moll), 1884, Rep. Voy. Chall., Zool. 9, p. 543, pi. cxiv, fig-17- Lenticulina ( Robulus ) altifrons Parr, 1950, B.A.N.Z.A.R. Exped. Rep. V, pt. 6, p. 323, pi. xi, figs. 12a, b. Five small specimens were found in sample 45. Brady’s figured specimen was from Paine Island, and Parr records the species from off Tasmania. ' (213) Lenticulina ( Robulus ) sp. cf. cultratus Montfort. cf. Robulus cultratus Montfort 1808, Conchyl. Syst. 1, p. 214, 54e genre ; Cushman and McCulloch, 1950, p. 296, pi. 37, figs. 7, 8. Specimens close to those figured by Cushman and McCulloch from the Western Pacific were common in sample 45. (214) Lenticulina ( Robulus ) formosus (Cushman). Cristellaria formosa Cushman, 1918 etc. (1923), U.S. Nat. Mus. Bull. 104, pt. 4, p. 110, pi. xxix, fig. 1, pi. xxx, fig. 6. One large specimen of this well-marked species was found in sample 45. (215) Lenticulina {Robulus) vortex (Fichtel and Moll). Nautilus vortex Fichtel and Moll, 1803, Test. Micr., p. 33, pi. ii, figs. d.i. Cristellaria vortex (Fichtel and Moll) of authors. Robulus vortex (Fichtel and Moll). Cushman, 1932 etc. (1933), p. 6, pi. 2, figs, la, b. One large specimen from sample 30, with strongly recurved, narrow chambers and lim- bate sutures is referred to this species. Family Polymoephinidae. Genus Guttulina d’Orbigny, 1839. (216) Guttulina regina (Brady, Parker and Jones). Polymorphina regina Brady, Parker and Jones, 1871, Trans. Linn. Soc. 27, p. 241, pi. 41, figs. 32a, b. Guttulina regina (Brady, Parker and Jones). Cushman and Ozawa, 1930, p. 34, pi. 6, figs. 1 and 2. FORAMJXIFERA — COLLINS 385 Two juvenile specimens from sample 45 are referable to this well-marked species, which is widely distributed in Australian waters. Genus Globulotubci n. gen. Test hyaline, perforate, circular in cross-section. triloculine " in chamber arrangement, later chambers successively further removed from the base, aperture radiate with internal, free, entosolenian tube curved back toward but not directly connected with the aperture of the preceding chamber. Genotype Globulotubci entosoleniformis n. sp. This genus is erected for the reception of species having the general arrangement of Globulina d'Orbigny, but having a free internal entosolenian tube. It is basically different from Siphoglobulina Parr in that the " siphon ” in the latter genus is within the wall and opens to the exterior at its proximal end. The existence of an entosolenian tube has been considered to be a character of generic value in the Lagenae ( vide Parr, 1947), and has also been made the distinguishing feature of the polymorphine genus Laryngosigma Loeblicli and Tappan. (217) Globulotubci entosoleniformis n. sp. (Plate IV, figs. 5a, b.) Species with the characters of the genus. Holotype from sample 45. Dimensions of holotype : Length 0-36 mm., diameter 0-22 mm. Rare in sample 45. The specimens show the feature described by Sidebottom in his Delos paper, in the absorption of the septal faces of earlier chambers, leaving the interior of the test as one clear space, the earlier chambers being marked only by their sutural junctions. Genus Pyrulina d’Orbigny. (218) Pyrulina extensa (Cushman). Polymorphina longicollis Brady (not of Karrer), 1884, Rep. Voy. Chall., Zool. 9, p. 572, pi. lxxiii, figs. 18 and 19. P. extensa Cushman, 1910 etc. (1913), U.S. Nat. Mus. Bull. 71, p. 90, pi. 41, figs. 1-3. Pyrulina extensa (Cushman). Cushman and Ozawa, 1930, p. 53, pi. 12, figs. 5a-c. One specimen of this deep-water species was found in sample 45. Genus Glandulina d’Orbigny. (219) Glandulina laevigata d’Orbigny. Nodosaria (Glandulina) laevigata d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 252, no. 1, pi. 10, figs. 1-3. This smooth form occurs in the samples of group 3 and is rather more common than other species of the genus. Both microspheric and megalospheric specimens were present. 386 GREAT BARRIER REEF EXPEDITION (220) Glandulina semistriata n. sp. (Plate IV, figs. 6 a, b.) Test fusiform, oral end bluntly rounded with a radiate aperture, apical end sharply pointed and surrounded by a very few small spines, surface finely striate except near the aperture, where it is smooth. The microspheric form (fig. 6) has 2 pairs of biserial chambers following the proloculus, then 2 uniserial, embracing chambers. The megalospheric form has a large proloculus followed by 3 uniserial embracing chambers. Holotype from sample 23. Dimensions of holotype : Length 0-34 mm., diameter 0-20 mm. The megalospheric form of this species is very close to Batsch’s fig. 2c of Nautilus ( Orthoceras ) comatus (1791, pi. 1), which has been identified by Cushman as Pseudoglandulina glans d’Orbigny, but the presence of the microspheric form establishes the species as a true Glandulina. Millett’s figure of Nodosaria ( Glandulina ) comata (Batsch), (1898 etc. (1902), pi. xi, fig. 2), is also referable to this species. (221) Glandulina echinata Millett. Nodosaria ( Glandulina ) echinata Millett, 1898 etc. (1902), Journ. Roy. Micr. Soc., p. 512, pi. xi, figs. 4 a, b. Two specimens from sample 23 and one from 45 agree with Millett’s figure in being ovate-fusiform with a short-lipped apertural neck and a development of short spines over the surface. An entosolenian tube is attached to the wall of the test from the aperture about half-way to the aboral end. Millett does not indicate the nature of the early chambers in his species, but they are definitely biserial in the present specimens. While, therefore, the identification with Millett’s species cannot be absolutely certain, it is considered to be likely, since the impor- tance of the arrangement of the early chambers separating the nodosarian Glandulinas (= Pseudoglandulina Cushman) from the polymorphine (= Glandulina s. str.) was not then recognized. Genus Sigmomorphina Cushman and Ozawa, 1928. (222) Sigmomorphina terquemiana (Fornasini). Polymorphina amygdaloides Reuss var. terquemiana Fornasini, 1902, Mem. Accad. Istut. Sci. Bologna, ser. 5, 9, p. 72, text-fig. 25. Sigmomorphina semitecta (Reuss) var. terquemiana (Fornasini). Cuslnnan and Ozawa, 1930, p. 129, pi. 33, figs. 4, 5. pi. 34, figs. 2, 3, pi. 35, fig. 1 ; Cushman, 1932 etc. (1933), p. 40, pi. 9, figs. 6-9. Two specimens from sample 45 are referable to this species, which has been recorded as Recent from the Mediterranean and the tropical Pacific. Since it has apparently been differentiated from S. semitecta s. str. since the Eocene, it appears best to treat it as a related but distinct species. Genus Laryngosigma Loeblich and Tappan. (223) Laryngosigma williamsoni (Terquem). Polymorphina williamsoni Terquem, 1878, Mem. Soc. Geol. France, ser. 3, 1, p. 37. Sigmomorphina williamsoni (Terquem). Cushman and Ozawa, 1930, p. 138, pi. 38, figs. 3, 4. Laryngosigma williamsoni (Terquem). Loeblich and Tappan, 1953, p. 84, pi. 16, fig. 1. Specimens from sample 45 are somewhat smaller but otherwise closely similar to specimens from the coast of England with which they were compared. The FORAMENTFERA— COLLINS 387 entosolenian tube distinguishing this genus from Sigmomorphina is clearly evident in these specimens. Superfamily BULIMINIDEA. Family Buliminidae. Subfamily Turrilixixae. Genus Buliminella Cushman, 1911. (224) Buliminella sp. cf. parallela Cushman and Parker. Bulimina elegant issim a Millett (not of d’Orbigny), 1898 etc. (1900), Journ. Roy. Micr. Soc., p. 276, pi. 2, fig. 4. cf. Buliminella parallela Cushman and Parker, 1931, Proc. U.S. Nat. Mus. 80, art. 3, p. 13, pi. 3, figs. 15a-c. A few specimens were found in most of the samples of groups 1, 2 and 3. They were of the type figured by Millett front the Malay Archipelago, questionably identified by Cushman and Parker as referable to their species B. parallela , the type of which was from the South Atlantic. The identification is probably correct, as there is sufficient variation in proportions and sutural depression in the Queensland specimens to include both figured specimens. (225) Buliminella latissima n. sp. (Plate IV, figs. 7 a, h, c.) cf. Bulimina elegantissima d’Orbigny var. compressa Millett (not B. compressa Bailey), 1898 etc. (1900), Journ. Roy. Micr. Soc., p. 277, pi. 2, fig. 5. Test irregularly ovate in outline, compressed, of 2-3 whorls with 5-6 chambers in the last whorl, chambers narrow with slightly depressed, limbate, sinuate sutures, spiral suture distinct, sutural margin of chambers lobed, aperture semicircular at the base of the flattened semicircular apertural face of the last chamber, with radiating striae. Holotype from sample 45. Dimensions of holotype: Length 0-26 mm., breadth 0-16 mm., breadth 0-16 mm., thickness 0-11 mm. Only two specimens of this form were found, but its compressed form and proportions are so distinctive that it was considered worthy of description as new. Millett’s var. com- pressa has many points of resemblance, the main difference being the more regular shape of the chambers and the absence of sutural lobing. Cushman and Parker (1947) have placed Millett’s form questionably in the synonymy of B. parallela, but the compressed form is a distinctive character, unusual in this genus, which justifies separation. The present specimens would have been referred provisionally to Millett’s variety had the name been available for use. In the circumstances it appears best to describe it as a new species. (226) Buliminella milletti Cushman. Buliminella milletti Cushman, 1933, Contr. Cushman Lab. Foram. Res. 9, pt. 4, p. 78, pi. 8, figs. 5, 6. Rare in the samples of groups 1 , 2 and 3. The species is recorded from many localities in the Indo-Pacific area and also from the West Indies. 388 GREAT BARRIER REEF EXPEDITION (227) Buliminella spicata Cushman and Parker. Buliminella madagascariensis (d’Orbigny) var. spicata Cushman and Parker (in Cushman), 1932 etc. (1942), U.S. Nat. Mus. Bull. 161, pt. 3, p. 8, pi. 3, figs. 5, 6. Frequent in samples 26 and 45. As this form has a similar geological range and occurs in the same geographical areas as B. madagascariensis s. str., it cannot be regarded as a subspecies, and is therefore here considered as a related but distinct species. Genus Buliminoides Cushman, 1911. (228) Buliminoides williamsonianus (Brady). Bulimina williamsoniana Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 56 ; 1884, p. 408, pi. li, figs. 16, 17. Buliminoides williamsonianus (Brady). Cushman, 1910 etc. (1911), U.S. Nat. Mus. Bull. 71, pt. 2, p. 90, text-fig. 144. This species is well distributed and common in the warmer waters of the Australian region, and is recorded by Parr (1945) from Bass Strait. Occasional in the samples of groups 1, 2 and 3. Genus Ungulatella Cushman, 1931. (229) Ungulatella pacifica Cushman. (Plate IV, fig. 8.) Ungulatella pacifica Cushman 1931, Contrib. Cushman Lab. Foram. Res. 7, p. 82, pi. 10, figs. 11, 12. Four specimens of this rare species were found in samples 29 and 30 from Jukes Reef, Outer Barrier. The species was described from the island of Rangiroa in the South Pacific, and it does not appear to have been recorded since then. Subfamily Bulimininae. Genus Bulimina d’Orbigny, 1826. (230) Bulimina barbata Cushman. Bulimina barbata Cushman, 1927, Scripps Inst. Oceanog. Bull., Tech. ser. 1, no. 10, p. 151, pi. 2, fig. 11. Specimens were common in sample 45. The species was described from the Western Pacific and has been recorded by Cushman from the Philippines. (231) Bulimina marginata d’Orbigny. Bulimina marginata d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 269, no. 4, pi. 12, figs. 10-12. This widely-distributed species was occasionally found in the samples of group 3 and was common in sample 45. (232) Bulimina rostrata Brady. Bulimina rostrata Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 408, pi. li, figs. 14, 15. Two specimens were found in sample 26. This species has a general Pacific distribution. FORAMIXIFERA— COLLIXS 389 (233) Bulimina oblonga n. sp. (Plate IV, figs. 9a, b , c.) Test short, thick, with both ends bluntly rounded, slightly increasing in width toward the oral end, in 2-3 whorls with 3 chambers to a whorl, sutures very slightly depressed and limbate, about 12 weak costae developed in the earlier part of the test, extending about half-way toward the oral end. Aperture loop-shaped, almost axial and commencing at the suture between the last 2 chambers. Holotype from sample 45. Dimensions of holotype : Length 0-25 nun., diameter 0-16 mm. Three specimens only were found, but the characters appear distinct enough for it to be described as new. It does not appear to be closely related to any described form. Genus Globobulimina Cushman, 1927, emend. Hoglund 1947. (234) Globobulimina australiensis n. sp. (Plate IV, figs. 10 a, b, c .) Test fusiform, both ends bluntly rounded, brownish in colour, wall transparent in later chambers, conspicuously perforate, chambers distinct, slightly inflated, the last extending about two-thirds of the length of the test proximally, chambers 3 to a whorl in the later part of the test. Aperture open, virguline, turning inwards without any collar, toothplate projecting slightly if at all, rolled into two-thirds of a circle, free edge smooth. Holotype from sample 45, where it is fairly common. Dimensions of holotype : Length 0-80 mm., diameter 0-42 mm. This species is very close to that described by Hoglund (1947) as Globobulimina sp. “ b ”, but the chambers are rather more inflated in adult specimens and the test wall is thick and translucent in the earlier chambers, transparent in the last one or two. Hoglund makes a point of the opacity of the wall in his specimens, but it is doubtful whether this can be considered to be a specific feature. All the specimens found were apparently megalo- spheric. Genus Virgulina d’Orbigny, 1826. (235) Virgulina complanata Egger. Virgulina schreibersiana Czjzek, var. complanata Egger, 1893, Abhandl. kon. bay. Akad. Wiss. Munchen, cl. ii, 18, p. 292, pi. 8, figs. 91, 92. V. complanata Egger. Cushman, 19376, p. 26, pi. 4, figs. 13-17. A compressed biserial form which occurs occasionally in the samples of all groups is referred to this species, which was described from off Western Australia and has been recorded from the Pacific islands, the Antarctic and south-eastern Australia. (236) Virgulina pauciloculata Brady. Virgulina pauciloculata Brady, 1884, Rep. Voy. Chalk, Zool. 9, p. 414, pi. lii, figs. 4 and 5. This species, distinguished by its elongate chambers extending proximally, is not uncommon in the samples of groups 3 and 4. It has a wide Pacific distribution. 390 GREAT BARRIER REEF EXPEDITION Subfamily Delosininae. Gfenus Delosina Wiesner, 1931. (237) Delosina complexa (Sidebottom). Polymorphina complexa Sidebottom, 1904, etc. (1907), Mem. Manchester Lit. Phil. Soc. LI, no. 9, p. 16, pi. iv, figs. 1-3 ; 1918, p. 145, pi. v, figs. 6, 7. Delosina complexa (Sidebottom). Earland, 1934, p. 127, pi. v, fig. 16. Two characteristic specimens were found in sample 36. The species has been recorded from the coast of New South Wales and from the Antarctic. Subfamily Reussellininae. Genus Reussella Galloway, 1933. (238) Reussella spinulosa (Reuss). Verneuilina spinulosa Reuss, 1850, Denkschr. Akad. Wiss. Wien, 1, p. 374, pi. 47, fig. 12. Reussella spinulosa (Reuss). Galloway, 1933, p. 360, pi. 33, fig. 4. Fairly common in all groups. The specimens are closely comparable with topotype material from Nussdorf, being broadly triangular in side view with short blunt spines on the basal edge of each chamber. (239) Reussella spinosissima (Costa). Verneuilina spinossisima Costa, 1856, Atti Acad. Pont. 7, pt. 2, p. 263, pi. 23, figs. 5 a-c. Reussella spinosissima (Costa). Cushman, 1937, p. 20 ; 1945, p. 38, pi. 7, figs. 1-3. A somewhat elongate form, with blunt spines at the basal angles of the chambers and a rather large, triangular aperture is referred to this species, which was described from the Pliocene of Italy and is recorded in the Recent state from the Mediterranean and the Red Sea. Rare in the samples of groups 1, 2 and 3. Genus Chrysalidinella Schubert, 1907. (240) Chrysalidinella dimorpha (Brady). Chrysalidina dimorpha Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 24 ; 1884, p. 388, pi. xlvi, figs. 20, 21 ; Millett, 1898 (1900), -p. 12, pi. 1, fig. 14 ; Heron-Alien and Earland, 1914 etc. (1915), p. 632, pi. xlvii, figs. 29-31. ' Chrysalidinella dimorpha (Brady). Schubert, 1907, p. 243. C. earlandi Cushman, 1945, p. 53, pi. 8, fig. 23. This species is fairly common in the samples of groups 1 , 2 and 3. Many of the specimens are of the elongate parallel-sided form figured by Heron- Allen and Earland (1915) and re-named by Cushman as C. earlandi. However, considerable variation in proportions is shown both in the present material and in a large series of specimens collected by the author from the coast near Dar-es-Salaam, East Africa. Both typical dimorpha and earlandi occur in the same material together with other variations in shape, and the separation of the latter form does not appear to be justified. FORAMINIFERA — COLLINS 391 Genus Mimosina Millett, 1900. (241) Mimosina affinis Millett. Mimosina affinis Millett, 1898 etc. (1900), Journ. Roy. Micr. Soc., p. 548, pi. iv, fig. 11 ; Heron-Alien and Earland, 1914 etc. (1915), p. 650 ; Cushman, 1945, p. 43, pi. 7, figs. 15, 16. This species was not uncommon in the samples of group 3. Specimens were close to Millett's figures in general form and surface characteristics, but the sutural aperture was in all cases a simple curved slit, not cribrate. However, Millett describes the aperture as “a slit, usually cribrate ”, so that this feature is apparently variable. (242) Mimosina rimosa Heron- Allen and Earland. (Plate IV, fig. 11.) Mimosina rimosa Heron-Alien and Earland, 1914 etc. (1915), Trans. Zool. Soc. London, 20, p. 650, pi. 1, figs. 5-11. Typical specimens were found in the samples of group 3. This species does not appear to have been recorded since its original description from the Kerimba Archipelago. (243) Mimosina echinata Heron- Allen and Earland. Mimosina echinata Heron-Alien and Earland. 1914 etc. (1915), Trans. Zool. Soc. London, 20, p. 651, pi. 50, figs. 12-18 ; Sidebottom; 1918, p. 128 ; Cushman, 1945, p. 44, pi. 7, figs. 20-22. Typical specimens were not uncommon in the samples of groups 3 and 4. The species is recorded by Sidebottom from off New South Wales. Genus Trimosina Cushman, 1927. (244) Trimosina milletti Cushman, subsp. multispinata, n. subsp. (Plate IV, fig. 12.) This species differs from T. milletti milletti in the development of a fringe of short spines on the lower margin of all the chambers, except the last 2 or 3. In the early part of the test the septation is largely obscured by this spinous development. The apertures are variable, that at the junction of the last and penultimate chambers varying from slit- like to arched, and the supplementary aperture in the distal face rounded to triangular. These differences, while constant, do not appear to justify specific distinction, and as the typical form does not occur in this material, the spinous variation is treated as a geographi- cal subspecies. Holotype from sample 44. Distribution generally in group 3, rare. Dimensions of holotype : Length 0-31 mm., greatest diameter 0-24 mm. Subfamily Uvigerininae. Genus Uvigerina d’Orbigny, 1826. (245) Uvigerina proboscidea Schwager. Uvigerina proboscidea Schwager, 1866, Novara-Exped., Geol. Theil. II, p. 250, pi. 7, fig. 96. U. proboscidea Schwager. Cushman and Todd, 1941, p. 73, pi. 17, fig. 9, pi. 19, figs. 3-9 ; Parr, 1950, p. 340. A few specimens were found in most of the samples of groups 1, 2 and 3. 392 GREAT BARRIER REEF EXPEDITION (246) Uvigerina porrecta Brady. Uvigerina porrecta Brady, 1879 etc., Quart. Journ Micr. Sci. XIX, n.s., p. 60, pi. viii, figs. 15, 16 ; 1884, p. 577, pi. lxxiv, figs. 22, 23. Rare in samples of group 3, rather more common in sample 45. Genus Siphouvigerina Parr, 1950. (247) Siphouvigerina fimbriata (Sidebottom). Uvigerina porrecta Brady, var. fimbriata Sidebottom, 1918, Journ. Roy. Micr. Soc. for 1918, p. 147, pi. v, fig. 23. Siphouvigerina fimbriata (Sidebottom). Parr, 1950, p. 342, pi. xii, fig. 22. This beautiful little species was found to be rare in samples 23, 24 and 45. Its recorded range is from off Java to the colder waters east of Tasmania. Genus Siphogenerina Schlumberger, 1883. (248) Siphogenerina striata curta Cushman. Sagrina striata Brady, 1884 (not of Scbwager), p. 584, pi. lxxv, figs. 25, 26. Siphogenerina striata (Scbwager), var. curta Cushman, 1926, Proc. U.S. Nat. Mus. 67, art. 25, p. 8, pi. 2, fig. 5, pi. 5, figs. 5, 6. Fairly common in samples of groups 1, 2 and 3. (249) Siphogenerina virgula (Brady). (Plate IV, figs. 13a, b, c.) Sagrina virgula Brady, 1879 etc., Quart. Journ. Micr. Sci. XIX, n.s., p. 61, pi. viii, figs. 19-21 ; 1884, p. 583, pi. lxxvi, figs. 4-7 (not 8-10). S. australiensis Goddard and Jensen, 1907, p. 294, pi. 6, figs. 3a-c. Siphogenerina virgula (Brady). Cushman, 1926, p. 14, pi. 2, figs. 7, 8, pi. 4, figs. 8, 9 (synonymy). Rectobolivina virgula (Brady). Hofker, 1951, p. 93, text-fig. 52. This species occurs in three forms in the present material. The megalospheric form has a proloculus nearly as big as the following uniserial chamber, and appears to be the rarest of the three forms. A microspheric form has a bolivine series of from 4 to 6 chambers, usually set at an angle to the later axis of growth, and another microspheric form has a reduced biserial portion of 2-3 chambers, practically circular in section and in line with the later axis. This latter form is by far the commonest and is met with in most of the samples. The author has topotype material of Goddard and Jensen’s Sagrina australiensis from the collection of the late W. J. Parr. The specimens are all of the microspheric form with reduced biserial portion referred to above, and lack the development of long spines pointing proximally from the apertural lip of each uniserial chamber. This is a common occurrence in shallow-water material and can be ascribed to erosion, as closely similar specimens from other shallow deposits show the full development of these delicate spines In the author’s opinion, S. australiensis as described by Goddard and Jensen is no more than a somewhat eroded microspheric specimen of Brady’s S. virgula. The present material bears out Hofker’s statement that the early portion of the test is biserial rather than uvigerine, as previously accepted. However, Bandy (1952) has shown that the type species of Siphogenerina, S. raphanus (Parker and Jones), includes forms FORAMLNTFERA — COLLIXS 393 with either biserial or triserial early chambers, so that S. virgula cannot be separated on this character. Rectobolivina is not available for this species, since by definition it includes only compressed forms such as the type, R. bifrons. The feature which appears to separate S. virgula from other Siphogenerinae is the absence of any connecting structure (siphon or toothplate) between successive apertures. The rudimentary toothplate observed by Hofker, which the present author has been unable to discern, hardly takes the place of the elaborate and well-developed structures seen in other species. It rather suggests the small tooth found in the genus Siphonodosaria, and it may be that the present species represents a stage in evolution between Siphogenerina and Siphonodosaria ; in fact, the megalospheric form in the absence of the others could well be referred to the latter genus. © Considering the present state of uncertainty in the classification of the species once included in Sagrina , it is perhaps best to retain this species in Siphogenerina pending further revision. Genus Trifarina Cushman, 1923. (250) Trifarina bradyi Cushman. Rhabdogonium tricarinatum Brady, 1884 (not Vaginulina tricarinata d’Orbigny), p. 525, pi. lxvii, figs. 1-3. Trifarina bradyi Cusliman, 1918 etc. (1923), p. 99, pi. xxii, figs. 3-9. Typical specimens were found in sample 45. Subfamily Gumbelininae. Genus Giimbelitria Cushman, 1933. (251) Giimbelitria vivans Cushman. (Plate IV, fig. 14.) Verneuilina pygmaea Millett, 1898 etc. (1900) (not of Egger), p. 11, pi. 1, figs. 13a, b. Giimbelitria ? vivans Cushman, 1934, Contrib. Cushman Lab. Foram. Res. 10, pt. 4, p. 105, pi. 13, figs. 9, 10. This rare species was not uncommon in the samples of group 3 and in sample 45. The specimens were all triserial throughout, none showing the irregular addition of the later chambers noted by Cushman. The first record of this form appears to be that of Millett (loc. cit) who considered it to be a hyaline isomorph of Verneuilina pygmaea Egger, and recorded it as “ scattered over the region ”. His Malay Archipelago material was mostly from shallow water, Cushman’s specimens were from off New Guinea in 129 fathoms, and the present specimens are from 600 metres and from shallow-water dredgings. As the characters of the species are substantially those of the Upper Cretaceous and Eocene genus Giimbelitria, it is here referred accordingly, though the absence of records of the genus from the later Tertiary suggests that there may not be a direct phylogenetic relationship, and that the Recent species may be a parallel development deriving from some other triserial form. Subfamily Plectofrondicul ariinae . Genus Bolivinella Cushman, 1927. (252) Bolivinella elegans Parr. Textularia folium Parker and Jones. Brady, 1884 (pars.), pi. xlii, figs. 3-5, (not 1, 2). Bolivinella elegans Parr, 1932, Proc. Roy. Soc. Victoria, 44, N.s., pt. II, p. 224. 394 GREAT BARRIER REEF EXPEDITION B. folia (Parker and Jones). Cushman, 1929 (not of Parker and Jones), p. 29, pi. 5, figs. 1,2; 1932 etc. (1942), p. 3, pi. 2, figs. 1-4, 6. B. folia (Parker and Jones), var. ornata Cushman. Cushman, 1932 etc. (1942) (not of Cushman, 1929), p. 5, pi. 2, figs. 5, 8. Considerable confusion has arisen between Bolivinella folium (Parker and Jones) and B. elegans Parr. It was shown by Parr (1932) that B. folium (typica) is a southern Australian species, characterized by broadly limbate sutures broken by fine diagonal striae which converge toward the distal end of the test, giving the effect of fine beading. This is the form described by Cushman (1929) as B. folia var. ornata, a name which should be relegated to the synonymy of B. folium. B. elegans is a species of the warmer waters of the Indo-Pacific, and is characterized by raised, thin ridge-like sutures without beading. Cushman (1942) records a single specimen of “ B. folia var. ornata ” from Levuka, Fiji, but illustrates two different specimens (figs. 5 and 8) which do not show the ornament of the southern Australian species, and in general fall within the rather wide limits of variation of B. elegans. This record is considered to be of doubtful authenticity. B. elegans is not uncommon in sample 45. Subfamily Bolivininae. Genus Bolivina d’Orbigny, 1839. (253) Bolivina abbreviata Heron-Alien and Earland. Bolivina limbata Brady, var. abbreviata Heron- Allen and Earland, 1924, Journ. Linn. Soc., Zool. 35, p. 622, pi. 36, figs. 25-27. B. abbreviata Heron-Alien and Earland. Cushman, 1937, p. 143, pi. 18, figs. 24, 25. Bare in the samples of groups 1, 2 and 3. Becords are all from the Pacific. (254) Bolivina alata Seguenza subsp. fimbriata, n. subsp. (Plate V, figs. 1 a, b.) This subspecies differs from B. alata alata in the keel being blunt and fimbriate, due to a row of tubules within it, rather than thin and imperforate as in the typical form. Otherwise the characters are those of the typical species, which was not found in the material. Holotype from sample 45, where it was fairly frequent. Dimensions of holotype ; Length 0-34 mm., breadth 0-17 mm., thickness 0-06 mm. (255) Bolivina compacta Sidebottom. Bolivina robusta Brady, var. compacta Sidebottom 1904 etc. (1905), Mem. Proc. Manchester Lit. Phil. Soc. 49, no. 5, p. 15, pi. 3. fig. 7. B. compacta Sidebottom. Cushman 1937, p. 135, pi. 17, figs. 22-24 (synonymy). Bare in the samples of groups 1, 2 and 3. (256) Bolivina hantkeniana Brady. Bolivina hantkeniana Brady 1879 etc. (1881), Quart. Journ. Micr. Sci. XXT, p. 58 ; 1884, p. 424, pi. lii, figs. 16-18; Cushman, 1937, p. 127, pi. 16, figs. 18-20 (synonymy). Bare in sample 45. FOE AMINTFERA — COLLINS 395 (257) Bolivina quadrilatera (Sell wager). Textularia quadrilatera Schwager, 1S66, Novara-Exped., Geol. Theil. II, p. 253, pi. 7, fig. 10. Bolivina quadrilatera (Schwager). Wright, 1891, p. 475 ; Hofker, 1951a. p. 102 et seq. Bolivinita quadrilatera (Schwager). Cushman, 1927, p. 90; 1932 etc. (1942), p. 2, pi. 1. As pointed out by Hofker, this species has all the characteristics of a typical Bolivina , and is not congeneric with the Cretaceous “ Bolivinita ” eleyi. Rare in sample 45. (258) Bolivina rhomboidalis (Millett). Textularia rhomboidalis Millett, 1898 etc. (1899), Journ. Roy. Micr. Soc., p. 559, pi. 7, fig. 4. Bolivina rhomboidalis (Millett). Cushman. 1922, p. 28 ; 1937, p. 138, pi. 18, fig. 7 (synonymy). Occasional in samples of groups 1, 2 and 3. (259) Bolivina subtenuis Cushman. Bolivina tenuis Brady (not of Marsson) 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 57; 1884, p. 419, pi. lii, fig. 29. B. subtenuis Cushman, 1936, p. 57, pi. 8, fig. 10 ; Collins, 1953, p. 101, pi. 1, fig. 7. Most of the records of this species are from the Fiji and Samoan islands, but it has been recorded by the author from Pleistocene deposits in western Victoria and is present in " Challenger " Stn. 185 material from Raine Island. Rare in sample 45. (260) Bolivina tortuosa Brady. Bolivina tortuosa Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 57; 1884, p. 420, pi. lii, figs. 31, 32 ; Cushman, 1937, p. 133, pi. 17, figs. 11-19 (synonymy). This typical Indo-Pacific species was common in the finer fractions of most of the samples. (261) Bolivina zanzibarica Cushman. Bolivina zanzibarica Cushman, 1936, Sp. Publ. no. 6, Cushman Lab. Foram. Research, p. 58, pi. 8, figs. 12a, b. This small species will probably be found to have an Indo-Pacific range similar to other East African species which occur in the Australian region. Rare in samples 24. Genus Loxostomum Ehrenberg, 1854. (262) Loxostomum convallarium (Millett). (Plate V, fig. 2.) Bolivina convallaria Millett, 1898 etc. (1900), Journ. Roy. Micr. Soc., p. 544, pi. lv, figs. 6a, b. Loxostoma convallarium (Millett). Cushman, 1937, p. 191, pi. 22, figs. 11-13. Hofker (1951) has referred this species to the genus Bitubulogenerina Howe, stating that the early chambers are triserial. No specimens with triserial early chambers were found in the present material, all being biserial throughout, with a tendency toward uniseriality in the later chambers. The specimens are very close to Millett’s original figures, which show smooth rounded chambers deeply undercut and crenulate at the base, features which are not apparent in Hofker’s figures of his specimens. The shell wall is smooth, polished 396 GREAT BARRIER REEF EXPEDITION and hyaline with scattered pores, not opaque as stated by Hofker ; nor do the chamber walls present nodular outgrowths as his figures indicate. It is therefore considered that the species is correctly placed in the genus Loxostomum. Specimens were not infrequent in the samples of groups 2 and 3. (263) Loxostomum strigosum (Brady). Bolivina lobata Brady, var. strigosa Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 425, pi. cxiii, fig. 7. Loxostoma strigosum (Brady). Cushman, 1937, p. 189, pi. 22, fig. 1 (synonymy). Two typical specimens of this rarely-recorded species were found in sample 45. (264) Loxostomum mayori (Cushman). Bolivina mayori Cushman, 1922, Publ. 311, Carnegie Inst. Washington, p. 27, pi. 3, figs. 5 and 6. Loxostoma mayori (Cushman). Cushman, 1937, p. 195, pi. 22, figs. 16-21 (synonymy). Bare in samples 32 and 45. It was described from the Tortugas and is widely recorded in the Pacific. (265) Loxostomum limbatum (Brady). Bolivina limbata Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 27 ; 1884, p 419, pi. lii, figs. 26-28. Loxostoma limbatum (Brady). Cushman, 1937, p. 186, pi. 21, figs. 36-39. This widely-distributed species was common in most of the samples. Genus Bifarina Parker and Jones, 1872. (266) Bifarina elongata Millett. (Plate V, fig. 3.) Bifarina elongata Millett, 1898 etc. (1900), Journ. Roy. Micr. Soc., p. 281, pi. 2, fig. 15. Two specimens from samples 38 and 45 appear to be referable to this species, which has apparently not been recorded since its original description. Its distinctive features are the triangular later chambers with spinous projections, the ornamentation with longitu- dinal rows of puncta, and the slit-like terminal aperture which takes up practically all of the distal end of the last chamber. The specimens are more regular in form than Millett’s figured specimen. (267) Bifarina queenslandica n. sp. (Plate V, figs. 4 a, b.) Test elongate, irregularly tapering, about 5 times the breadth in length, compressed oval in cross-section, chambers biserial in early stages with the axis of growth somewhat curved or twisting, becoming straight in later portion, last two or three chambers uniserial. Periphery rounded, slightly lobulate, sutures limbate and slightly depressed, chamber wall coarsely punctate giving a reticulate and semi-opaque appearance. Aperture terminal, within a roughly oval depression into which on one side projects a tooth-like extension of the chamber wall. The actual aperture is obscure, apparently slit-like. In broken speci- mens the apertural foramen is seen to be an oval hole surrounded by a raised rim or low collar, but there is no indication of any connecting structure between successive apertures. Holotype from sample 1 1 . FORAMINTFERA — COLLIXS 397 Dimensions of holotype : Length 0-57 mm., breadth 0-16 nun., thickness 0-11 mm. Rare in the samples of groups 1. 2 and 3. seldom more than one or two specimens in a sample. This species does not seem to be closely related to any described form. Genus Rectobolivina Cushman, 1927. (268) Rectobolivina bifrons (Brady). Sagrina bifrons Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 64 ; 1884, p. 582, pi. Ixxii, figs. 18-20. Rectobolivina bifrons (Brady). Cusliman, 1927a, p. 68, pi. 14, fig. 11 ; 1937, p. 204, pi. 23, figs. 13 and 14 (synonymy). Two specimens were foimd in sample 45. Family Cassidullnidae. Genus Cassidulina d'Orbigny, 1826. (269) Cassidulina elongata Sidebottom. Cassidulina bradyi Norman, var. elongata Sidebottom, 1904 etc. (1905), Mem. Proc. Manchester Lit. Phil. Soc. 49, no. 5, p. 17, pi. 3, fig. 11 ; Heron-Alien and Earland, 1914 etc. (1915), p. 653, pi. 1, fig. 20. One specimen of this widely-distributed but seldom recorded form was found in sample 30. It appears to have as wide a range of occurrence as C. bradyi s. str., from the NW. Atlantic to the Indo-Pacific, so that it is best considered as a separate species. (270) Cassidulina laevigata d'Orbigny. Cassidulina laevigata d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 282, pi. xv, figs. 4 and 5 ; Brady, 1884, p. 428, pi. liv, figs. 1-3. Four specimens of this wide-ranging species were found in sample 32. Genus Orthoplecta Brady, 1884. (271) Orthoplecta clavata Brady. Cassidulina ( Orthoplecta ) clavata Brady, 1884, Rep. Yoy. Chall., Zool. 9, p. 432, pi. cxiii, fig. 9. One specimen was found in sample 32. The species was described from the Admiralty Islands. Family Chilostomellidae. Genus Chilostomella Reuss, 1850. (272) Chilostomella oolina Schwager. Chilostomella oolina Schwager, 1878, Bull. Com. Geol. Ital. 9, p. 527, pi. 1, fig. 16 ; Cushman, 1926a, p. 74, pi. 11, figs. 3-10. Three small, elongate, thin- walled specimens from sample 45 are referred to this species. 398 GREAT BARRIER REEF EXPEDITION Genus Seabrookia Brady, 1890. (273) Seabrookia pellucida Brady. Seabrookia pellucida Brady, 1890, Journ. Roy. Micr. Soc., p. 567, text-fig. 60, la-c ; Millett, 1898 etc. (1901), p. 3, pi. 1, figs. 4 a-d ; Chapman, 1910, p. 406 ; Heron-Alien and Earland, 1922, p. 141 ; Parr, 1950, p. 346. Not uncommon in sample 45. Occasional specimens were found in the samples of group 3. Genus Pullenia Parker and Jones, 1862. (274) Pullenia quinqueloba (Reuss). Nonionina quinqueloba Reuss, 1851, Zeitschr. deutsch. Geol. Ges. 3, p. 71, pi. 5, fig. 31. Pullenia quinqueloba (Reuss). Brady, 1882, xi, p. 712 ; Cushman and Todd, 1943, p. 10, pi. 2, fig. 5 ; pi. 3, fig. 8 (synonymy). Three specimens were found in sample 45. The two smaller ones had 5 chambers in the final whorl, the larger had 4 but they are considered to be conspecific. Genus Sphaeroidina d’Orbigny, 1826. (275) Sphaeroidina bulloides d’Orbigny. Sphaeroidina bulloides d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 267, no. 1 ; Brady, 1884, p. 620, pi. lxxxiv, figs. 1 to 5 (not 6, 7). Common in sample 45, occasional in the samples of group 3. Genus Nonionella Cushman, 1926. (276) Nonionella pulchella Hada. Nonionella pulchella Hada, 1931, Tohoku Imp. Univ. Sci. Repts., ser. 4, Biol. 6, p. 120, Text-fig. 79. Rare in sample 24 (2 specimens). This species was described from Mutsu Bay, Japan. Genus Nonion Montfort, 1808. (277) Nonion subturgidus (Cushman). Nonionina subturgida Cushman, 1924, Publ. 342, Carnegie Inst. Washington, p. 47, pi. 16, fig. 2. Nonion subturgidum (Cushman). Cushman, 1932 etc. (1933), p. 43, pi. 10, figs. 4-7. Not uncommon in the samples of group 3, many specimens in sample 27. (278) Nonion cf.' depressulus (Walker and Jacob). cf. Nautilus depressulus Walker and Jacob, 1798, in Adams’ Essays (Kanmacher’s Edn.) p. 64, pi. 14, fig. 33. cf. Nonion depressulus (Walker and Jacob). Cushman, 1939, p. 20, pi. 5, figs. 22-25. Occasional specimens in the samples of groups 1, 2 and 3 are doubtfully referred to this species. (279) Nonion scapha (Fichtel and Moll). Nautilus scapha Fichtel and Moll, 1798, Test. Micr., p. 105, pi. 19, figs. d-f. Nonion scaphum (Fichtel and Moll). Cushman, 1939, p. 20, pi. 5, figs. 18—21 (synonymy). Rare, one specimen being found in sample 9. FORAMTXIFERA — COLLIXS 399 Super family ROTALIDEA. Family Spirillixidae. Subfamily Spirillixixae. Genus Spirillina Ehrenberg, 1843. (280) Spirillina vivipara Elirenberg. Spirillina vivipara Ehrenberg, 1843, K. Akad. Wiss. Berlin, Physik Abh., p. 442, pi. iii, sec. 7, fig. 41. This widely-distributed species is frequent in the finer fractions of most of the samples. (281) Spirillina limbata Brady. Spirillina limbata Brady, 1879 etc., Quart. Journ. Micr. Sci. XIX, N.S., p. 278, pi. viii, figs. 6a, b ; 1884, p. 632, pi. lxxxv, figs. 18-21 ; Parr, 1950, p. 348. Rare in sample 29 (1 specimen). (282) Spirillina decorata Brady. Spirillina decorata Brady, 1884, Rep. Yoy. Chalk, Zool. 9, p. 632, pi. lxxxv, figs. 22-25. Rare in sample 45 (1 specimen). (283) Spirillina inaequalis Brady. Spirillina inaequalis Brady, 1879, Quart. Journ. Micr. Sci. XIX, p. 278, pi. viii, figs. 25a, b; 1884, p. 631, pk lxxxv, figs. 8-11. Occasionally found in samples of group 3 with S. vivipara, but less common than that species. Genus Planispirillina Bermudez, 1952. (284) Planispirillina denticulata (Brady). Spirillina limbata Brady, var. denticulata Brady, 1884, Rep. Voy. Chalk, Zook 9, p. 632, pk lxxxv, fig. 17. S. denticulogranulata Chapman, 1909, p. 133, pk x, fig. 6. S. denticulata Brady. Ovey, 1947, pp. 17, 18 ; Parr, 1950, p. 351. The recorded range of this species is from Raine Island to Bass Strait. It was found to be rare in the samples of group 3. Genus Mychostomina Berthelin, 1881. (285) Mychostomina revertens (Rhumbler). Spirillina vivipara Ehrenberg, var. revertens Rhumbler, 1906, Zook Jahrb. Abt. Syst. Geol. Biol 24, p. 32, pk 2, figs. 8-10. Mychostomina revertens (Rhumbler). Galloway, 1933, p. 88, pk 7, figs. 6, 7 ; Bermudez, 1952, p. 27, pk 1, figs. 4—6. Occasional specimens were found in the finer fractions of samples of all groups vi. 6. 27 400 GREAT BARRIER REEF EXPEDITION Subfamily Patellinae. This subfamily is usually considered as containing four genera, Patellina, Patellinella, Annulopatellina and Patellinoides. Hofker (1951) has recently combined three of them ( Patellina , Patellinella and Patellinoides) in a new genus Discobolivina, which is obviously invalid, since the only name available for a combination of these three genera is that of the first-described, Patellina. His reasons for this combination are based on the morphology of the aperture and associated structures and on the pore system, and he considers that the many records which refer to an undivided spiral commencement in these genera are erroneous (p. 422, loc. cit.). This contention can scarcely be upheld against the weight of contrary evidence, particularly the detailed researches of Myers (1935) in the case of Patellina. In the present collections there are specimens of Patellina corrugata Williamson in which the earlier whorls are clearly undivided and spirilline. The author can, moreover, vouch for the occurrence of an undivided spiral commencement in the Tertiary species described as Patellinella annectens Parr and Collins, 1930. The genus Patellinoides, described fron the Antarctic by Heron-Alien and Earland (1934), is characterized by undivided early whorls, as shown clearly in the figures and also described. In contrast, the type species of the genus Patellinella (P. inconspicua Brady) which is common in the shore sands of the region from which it was described, Bass Strait, is either pseudobiserial from the beginning or in some specimens commences with a short trochospiral series of 2 or 3 chambers. In such specimens the early chambers may show a brown coloration similar to that found in Discorbis and other rotaline genera. The four species of “ Discobolivina ” illustrated by Hofker (loc. cit.) also start with a short trocho- spiral series. The investigations of Wood (1948) on the structure of the foraminiferal test show that Spirillina, Patellina and Patellinoides tend to have the test made up of a single crystal of calcite, whereas Patellinella is shown to have a radial wall structure similar to that of most other Rotalidae. These considerations suggest that two distinct lineages exist in this subfamily. Patel- lina and Patellinoides are characterized by having an undivided spiral commencement and single-crystal wall structure, suggesting that these genera have evolved from a spirilline ancestor, while Patellinella has a trochospiral early series and radial wall structure, suggest- ing discorbine ancestry. It follows that the pseudobiserial growth-plan of the later chambers in the three patellinid genera, actually a trochospiral coil with the chambers added at 180° intervals, as shown by the alternating slant of 'the loop-shaped aperture, is a case of convergence, producing similar apertural structures in genera which are not closely related. The tertiary species Patellinella annectens should probably be referred to the genus Patellinoides on account of its spirilline commencement. The position of this species and of the genus Annulopatellina Parr and Collins requires detailed examination of their wall structure, which is beyond the scope of this report. In the light of the above considerations, the genera Patellina and Patellinella have been retained in this work, with Discobolivina placed in synonymy, and Patellinella transferred to the subfamily Discorbisinae. Parr (1950) appears to have arrived at the same conclusion by including Patellinella in the Discorbidae without comment. FORA^IIXIFERA— COLLIXS 401 Genus Patettina 'Williamson, 1858. (286) Patettina corrugata Williamson. Patettina corrugata Williamson, 1858, Rec. For. Gt. Britain, p. 46, pi. iii, figs. 86-89 ; Parr and Collins, 1930, p. 90, pi. iv, figs. 1-5. The specimens found in this material are indistinguishable from the form found in British waters. Occasional specimens were found in samples of all groups. (287) Patettina altiformis Cushman. Patettina advena Cushman, var. altiformis Cushman, 1933a, Contrib. Cushman Lab. Foram. Res. 9, pt. 4, p. 87, pi. 9, figs. 8a, b. P. altiformis Cushman. Parr, 1950, p. 352. Two specimens of this high-spired form were found in sample 29, Jukes Beef. It is present in Raine Island material but, strangely enough, is not found in the deeper-water sample 45, where it is replaced by the wider-ranging P. corrugata. Parr records it from off Tasmania. Family Rotalidae. Subfamily Discorbisixae. Genus Discorbis Lamarck, 1804. (288) Discorbis rugosus (d’Orbigny). Rosalina rugosa d’Orbigny, 1839, Foram. Amer. Mend., p. 42, pi. ii, figs. 12-14. Discorbina rugosa (d’Orbigny). Brady, 1884, p. 652, pi. Ixxxvii, figs. 3a-c ; Millett, 1898 etc. (1903), p. 703 ; Chapman, 1901, p. 190 ; Heron-Alien and Earland, 1914 etc. (1915), p. 697. The umbilical flaps developed on the ventral side of the test in this species, though simple in form, are apparently homologous with the more complex structures found in D. vesicularis and related forms. It is therefore considered as a species of Discorbis s. str. Common in sample 29, occasional in other samples of group 3. (289) Discorbis subvesicularis n. sp. (Plate V, figs. 5a, b and c .) Test trochoid, planoconvex, periphery broadly rounded and lobulate, 2-3 whorls, 6 chambers in the last whorl, wall thin and coarsely perforate. Dorsal side convex, chambers inflated, sutures depressed and recurved, in some specimens having a small triangular opening between the chambers at the junction of the suture with the previous whorl, passing between the chamber walls to the opposite side of the test. Ventral face slightly convex, sutures radial and slightly recurved, umbilicus masked by strongly-developed extensions of the chamber wall forming umbilical flaps. Commencing at the junction of the apertural face with the previous whorl, the edge of the flap is everted to form a broad lip covering a crescentic aperture extending radially toward the umbilicus. From the end of this primary aperture the edge of the flap turns at right angles and passes over and is attached to the surface of flaps formed by previous whorls. Where it meets the last suture, a short narrow crescentic lip is formed, covering an accessory aperture. From the end of this aperture the flap is attached to the wall of the preceding chamber, and then turns distally, parallel to the direction of coiling, and forms a longer, sigmoid lip covering a second accessory 402 GREAT BARRIER REEF EXPEDITION aperture. This aperture is terminated by the incurving and merging of the edge of the flap with the chamber wall. Each flap, therefore, forms 3 arched, lipped apertures, separated by attached parts of the margin. The primary aperture is covered by the flap of the next succeeding chamber, the others remaining open in close proximity to the accessory apertures of the previous and following chambers. The flaps are perforate, but the pores are smaller and more scattered than those of the chamber wall proper. While the apertural flaps do not form separate chamberlets, partial separation is made by a fold or partition about one-third the height of the chamber, originating from the peripheral end of the primary aperture, attached to the dorsal interior surface of the chambers for about half its length, running in a sigmoid curve and in a roughly circumferential direc- tion to terminate on the ventral wall of the chamber just below the second accessory aper- ture. Immediately below and behind this point of attachment is the septal foramen con- necting to the penultimate chamber (Text-fig. 1). Text-fig. 1. — Discorbis subvesicularis n. sp. Specimen broken to show internal structures, af apertural foramen, aa accessory apertures, p partition, pa primary aperture, uf umbilical flap. Earlier chambers have the same flap-like structure but simplified in form. The whole complex of cavities, baffles and passageways formed in the umbilical area by these flaps and the sigmoid partitions invites comparison with the canal systems of the more highly- developed rotalids. No attempt has been made to apply the morphological nomenclature introduced by Hofker in describing these structures, as it is considered that the assumption of homologies which are not yet fully investigated is not advisable at this stage. Holotype from sample 1 . Dimensions of holotype ; diameter 0-51 mm., thickness 0-22 mm. A few specimens were found in most of the samples of groups 1, 2 and 3. This species is one of the vesicularis-dimidiatus group. It differs from D. vesicular is Lamarck from the Eocene of the Paris Basin in its rounded periphery ( D . vesicularis being keeled, at least in the earlier chambers) and its more complex development of the umbilical region. It differs from the southern Australian 'recent species, D. dimidiatus (Parker and Jones), in its consistently smaller size, rounded periphery and somewhat less complicated umbilical structures. Cushman (1918 etc. (1931), p. 34) refers to a speciesfrom the Australian coast “ which is very close to the Eocene species of Lamarck ”, probably the form under con- sideration. The closely allied species D. dimidiatus has been made the genotype of a new genus, Lamellodiscorbis Bermudez, 1952. After comparing it with specimens of D. vesicularis from the Eocene of Grignon, the author is of the opinion that the umbilical structures in the Australian species are no more than a larger and more complicated development of the alar projections seen in D. vesicularis, which are in some specimens widened at the FORAMINIFERA — COLLINS 403 extremities and recurved to form rudimentary supplementary apertures of similar nature to those foimd in the present species and in D. dimidiatus, though less developed. The differences between the present species and D. vesicularis are considered to be of specific value onlv. and it is therefore retained in Discorbis s. str. J 7 In describing this species I have had the privilege of consulting an unpublished MS. dealing with the morphology of D. dimidiatus. by Mr. A. N. Carter of the Department of Mines, Victoria, to whom my thanks are due for the assistance thus given in working out the somewhat similar structures in the present species. Genus Discopulvinulina Hofker, 1951. (290) Discopulvinulina bertheloti (d’Orbigny). Rosalinabertheloti d’Orbigny, 1839, in Barker- Webb and Berthelot, Hist. Nat. lies Canaries, 2, pt. 2, “ Foraminiferes,” p. 135, pi. 1, figs. 28, 30. Discorbis bertheloti (d’Orbigny). Cushman, 1918 etc. (1931), p 16, pi iii, fig. 2. Discopulvinulina bertheloti (d’Orbigny). Hofker, 1951, p. 449. Rare in the samples of group 3. (291) Discopulvinulina mira (Cushman). Discorbis mira Cushman, 1922. Publ. 311, Carn. Instit. Washington, p. 39, pi. 6, figs. 10, 11 ; 1918 etc. (1931), p. 25, pi. 5, figs. 5, 6a-c ; Parr, 1943, p. 16 ; Collins, 1953, p. 103. Described from the West Indies, this species is evidently well distributed in Australian waters. It was recorded from Glenelg, S. Australia by Parr (1943), and from warm-water Pleistocene deposits in Victoria by the author (1953). Specimens are not uncommon in the samples of groups 1, 2 and 3, and compare well with West Indian examples. (292) Discopulvinulina lobatula (Parr.) Discorbina isabelleana Brady (not Rosalina isabelleana d’Orbigny) 1884, Rep. Voy. Chall., Zool. 9, p. 646, pi. Ixxxviii, fig. 1. Discorbis lobatulus Parr, 1950, p. 354, pi. xiii, figs. 23-25. A fewT examples of this species were found in sample 45. It has a wide range in the Indo- Pacific and was recorded by Parr from off Tasmania. (293) Discopulvinulina subcomplanata (Parr). Discorbis subcomplanatus Parr, 1950, B.A.N.Z.A.R. Exped. Rept. V, pt. 6, p. 355, pi. xiv, figs, la, b, 2. Rare in sample 1. This species was described from the Antarctic and is distinguished by its heavily limbate sutures on the dorsal side and its complanate form. Genus Rosalina d’Orbigny, 1839. (294) Rosalina orbicularis Terquem. Rosalina orbicularis Terquem, 1876, Anim. sur la plage Dunkerque, p. 75, pi. ix, figs. 4a, b. Discorbis orbicularis (Terquem). Berthelin, 1878, p. 39, no. 63 : Cushman, 1918 etc. (1931), p. 27, pi. 6, figs. 3a-c (synonymy). Rosalina orbicularis Terquem. Bermudez, 1952, p. 34. Good examples are common in samples 44 and 45. 404 GREAT BARRIER REEF EXPEDITION (295) Rosalina orientalis (Cushman). Discorbis globularis Heron-Alien and Earland (not of d’Orbigny), 1914 etc. (1915), p. 694, pi. li, figs. 36-39. D. orientalis Cushman, 1925, Bernice P. Bishop Mus., Bull. 27, p. 130. Probably the commonest of the smaller rotalids in the samples of groups 1, 2 and 3, this species has a wide Indo-Pacific range. It was described as having “ the umbilical end of each chamber ending in a tooth-like process, bluntly pointed These processes originate in a flap-like extension of the chamber wall covering the umbilical extension of the aperture. As chambers are added, thickening of the shell wall at this point results in the formation of polished non-tubulate blebs on the inner margin of the chambers, varying from triangular tooth-shaped inwardly-pointing processes to groups of irregularly-rounded pustules. As this form is related to Rosalina globularis d’Orbigny, it should be referred to that genus. (296) Rosalina frustata (Cushman). Discorbis frustata Cusliman, 1933, Contrib. Cushman Lab. Foram. Res. 9, pt. 4, p. 88, pi. 9, figs. 2 a-c. Described by Cushman from off Fiji, this species is very distinctive with its tiers of upward-curved, flanged chambers. Two specimens were found in sample 29. Genus Mississippina Howe, 1930. (297) Mississippina pacijica Parr. Mississippina pacijica Parr, 1950, B.A.N.Z.A.R. Expect Rept. V, pt. 6, p. 361, pi. xiv, figs. 17a-c. Five specimens were found in sample 45. Parr records the species from Raine Island. The state of preservation of these specimens apparently differs from that of Parr’s material from off Tasmania, the body of the test being white and opaque, while the peripheral areas are yellowish and translucent with a depressed margin. Genus Conorbella Hofker, 1951. (298) Conorbella patelliformis (Brady). Discorbina patelliformis Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 647, pi. lxxxviii, fig. 1 ; Millett, 1898 etc. (1903), p. 700 ; Heron-Alien and Earland, 1914 etc. (1915), p. 703, pi. Hi, fig. 32 ; Sidebottom, 1918, p. 254. Discorbis patelliformis (Brady). Cushman, 1924, p. 33. Conorbella patelliformis (Brady). Bermudez, 1952, p. 37. This well-known Indo-Pacific species was found occasionally in the samples of groups 1, 2 and 3. (299) Conorbella pulvinata (Brady). Discorbina pulvinata Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 650, pi. lxxxviii, figs. 10a, b ; Heron-Alien and Earland 1914 etc. (1915), p. 702. Conorbella pulvinata (Brady). Hofker, 1951, p. 470. Rare, two specimens found in sample 26. FOBAMTNIFERA— COLLINS 405 (300) Conorbella pyramidalis (Heron-Alien and Earland). Discorbina pyramidalis Heron-Alien and Earland, 1924, Journ. Linn. Soc. Zool. 35, p. 634, pi. 37, figs. 56-61. Conorbella pyramidalis (Heron-Alien and Earland). Hofker, 1951, p. 470. This species, described from Lord Howe Island and recorded by the authors from South and Western Australia, is represented by rare but typical specimens in the samples of groups 3 and 4. (301) Conorbella tabernacidar is (Brady). Discorbina tabernacularis Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 648, pi. lxxxix, fig. 7 (not 5 and 6). Conorbella tabernacularis (Brady). Hofker, 1951, p. 470. Striate specimens similar to Brady’s fig. 7 are not uncommon in all groups. It is doubtful, however, whether this form is conspecific with the large costate form of figs. 5 and 6, which was not found in the present material. (302) Conorbella corrugata (Millett). Discorbina corrugata Millett, 1898 etc. (1903), Journ. Roy. Micr. Soc., p. 700, pi. vii, fig. 5. Rare in the samples of group 3, less frequent than C. pyramidalis, from which it differs in the greater number (5 to G) and less regularity of the radial ridges on the dorsal side. (303) Conorbella opercularis (d’Orbigny). Rosalina opercularis d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 271, no. 7. Discorbina opercularis (d’Orbigny). Brady, 1884, p. 650, pi. lxxxix, figs. 8, 9. Conorbella opercularis (d’Orbigny). Bermudez, 1952, p. 37. Four specimens from sample 45 are similar to those figured by Brady, which were from Queensland waters. Parr (1950) records the species from off Tasmania. (304) Conorbella earlandi n. sp. (Plate V, figs. 6u, b and c .) Test minute, compressed, subhexagonal in outline, with slightly inflated chambers and a narrow keel. On the ventral side the sutures are straight and radial, slightly depressed, chambers triangular and somewhat inflated, with a depressed umbilicus. On the dorsal side all chambers are visible, subrectangular, 3 to 4 times as long as wide, arranged in a spiral hexagonal series, surface concave with the ends of the chambers curving up to form ridges which are continuous from the centre of the test to the periphery. Ventral side smooth, very finely perforate, dorsal side more coarsely perforate. Aperture is a narrow slit at the anterior margin of the last chamber on the ventral side, extending from the periphery to the umbilicus. Holotype from sample 45. Dimensions of holotype : Diameter 0-20 mm., thickness 0-07 mm. Three specimens were found in sample 45, but a good series was available fromRaine Island material. The geometric outline and ridged dorsal side suggest relationships with C. pyramidalis and C. corrugata , but in this case the ridges are caused by the upward- curved radial sutures, not by an angular form of the chamber itself. The species does not appear to be closely related to any described form. 406 GREAT BARRIER REEF EXPEDITION This species was recognized as new by Mr. A. Earland in a slide mounted by him from Eaine Island material and given to the late W. J. Parr for further study. I have great pleasure in associating it with his name, not only for this reason, but in inadequate acknow- ledgment of his helpful advice and encouragement over a period of many years. Genus Discorinopsis Cole, 1941. (305) Discorinopsis tropica n. sp. (Plate V, figs. 7 a, b and c.) Test trochospiral in about 2 whorls, up to 10 chambers in last whorl ; dorsal side rounded with all chambers visible, coarsely perforate, sutures depressed-limbate ; ventral side flat or slightly concave, all but the last 2 chambers masked by irregular pustular shelly outgrowths. Periphery bluntly keeled, aperture a low slit on the ventral side of the periphery. Holotype from sample 9. Dimensions of holotype : Greater diameter 0-52 mm., thickness 0-17 mm. One specimen was found in sample 9, and two others were available for study from the collection of the late W. J. Parr, their locality being Plum Beach, New Caledonia. The species differs from D. aguayoi Bermudez in its more regular form and bluntly keeled periphery. Genus Bronnimannia Bermudez, 1952. (306) Bronnimannia haliotis (Heron-Alien and Earland). Discorbina haliotis Heron- Allen and Earland, 1924, Journ. Roy. Micr. Soc., pt. 2, p. 173, pi. xiii, figs. 99-101. Discorbis haliotis (Heron-Alien and Earland). Parr, 1939, p. 68 ; Collins, 1953, p. 96. Planulina haliotis (Heron-Alien and Earland). Chapman, 1941, p. 176. This species was described from the Lower Miocene of Victoria and has since been recorded from the Lower Pliocene and Pleistocene. Parr records it from Geraldton Harbour, Western Australia, and mentions it as occurring at Barwon Heads, Victoria. Chapman records the species from 470 fathoms off the south-east coast of Australia. It occurs occasionally in the samples of groups 3 and 4. Heron- Allen and Earland’ s figure (loc. cit.) does not show the full characteristics of the species. The sutures on the dorsal side are usually incurved, varying from a shallow inden- tation to a deep re-entrant, and leaving the umbilical margin of the chamber free of attachment. The author’s figure shows a faint suggestion of this form, but gives a wrong impression of the suture as a simple curve. Specimens from the type locality show all stages, those with a deep re-entrant sinus being the more common. There is considerable variation in thickness of the test in the present material, some specimens being flat and compressed, others intermediate between these and the typical form. This species is very close to the genotype, B. palmerae (Bermudez) which, however, is considerably thicker in proportion to diameter. Genus Poroeponides Cushman, 1944. (307) Poroeponides lateralis (Terquem). Rosalina lateralis Terquem, 1878, Mem. Soc. Geol. France, ser. 5, 1, mem. Ill, p. 25, pi. ii, figs. 11a, b. Poroeponides lateralis (Terquem). Cusliman, 1944, p. 34 (synonymy). FORAMTNIFERA — COLLINS 407 This common Indo-Paeific species was met with in most of the samples, particularly those of group 3. Genus Patellinella Cushman, 1928. (308) Patellinella jugosa (Brady). Textularia jugosa Brady, 1884, Rep. Yoy. Chall., Zool. 9, p. 358, pi. xlii, figs, la, b. T. inconspicua Brady var. jugosa Brady. Millett, 1898 etc. (1899), p. 558, pi. vii, figs. 2 a, b. Discobolivina inconspicua (Brady). Hofker, (not of Brady) 1951, p. 431, text-fig. 296. This well-marked species with ridged projecting sutures was rare in sample 45, one specimen only being found, which was closely comparable with specimens from Raine Island, the type locality. It is cpiite distinct from Brady’s Textularia inconspicua which, as figured by Brady (1884) and by Parr and Collins (1930), has a smooth subconical outline with wide flush sutures of clear shell substance, the chamber walls being perforate with large closely-spaced pores. The latter species is common in Bass Strait, from which locality it was described. It has not been met with in the present samples, nor, in the author’s experi- ence, in any tropical material, though Brady records it from Nares Harbour, Admiralty Islands and from south of Japan, and Heron- Allen and Earland record it from the Kerimba Archipelago. The latter record does not appear to be correct, as the figured specimen has thin depressed sutures. Millett’s figure of Textularia inconspicua (1898 etc., (1899), pi. vii, fig. 1). with its flared outline and markedly convex apertural face, is also distinct from the typical southern Australian form. P. jugosa appears to be confined to tropical Australia and the regions to the north. (309) Patellinella nitida (Hofker). Discobolivina nitida Hofker, 1951, Siboga Expect, Part III, p. 498, text-figs. 293, 294. This small depressed form, subcircular in end view, corresponds well with Hofker’s description of specimens from south of Sumatra. Similar specimens were present in material from Challenger Stn. 185, Raine Island, mounted by Mr. A. Earland. Rare in sample 45. (310) Patellinella carinata n. sp. (Plate V, figs. 8 a, b, c.) Test minute, hyaline, commencing with a short trochoid spiral of 3 or 4 chambers, with later chambers added at 180° intervals (Text-fig. 2). Triangular in side view, with periphery slightly carinate, and set with fine apically-directed spines. Sutures distinct, Text-fig. 2. — Patellinella carinata n. sp. Specimen showing arrangement of early chambers. arcuate, thin and almost flush, produced at the periphery as a larger spine. Chamber wall thin, transparent, finely perforate, with an irregular row of larger pores on the proximal side of each suture. In apertural view the test is ovoid with pointed ends, the distal chamber wall being thin, transparent and finely perforate, with a few larger pores scattered round the periphery and grouped near the outer ends. Aperture loop-shaped with a curved 408 GREAT BARRIER REEF EXPEDITION toothplate, the aperture of the penultimate chamber being closed by what appears to be an extension of the toothplate. Holotype from sample 45. Dimensions of holotype : Length 0-18 mm., breadth 0-28 mm., thickness 0-16 mm. This species is distinguished from others of the genus by its carinate and spinous periphery. It is fairly rare in the deeper-water sample 45, and is present in material from Challenger Stn. 185, Raine Island. Subfamily Valvulineriinae. Genus Globorotalia Cushman, 1927. (311) Globorotalia tumida (Brady). Pulvinulina menardii var. tumida Brady, 1877, Geol. Mag., Dec. II, 4, p. 535 ; 1884, p. 692, pi. ciii, figs. 4—6. Globorotalia tumida (Brady). Cushman, 1927, p. 175. Rare in sample 45 (one specimen). (312) Globorotalia canariensis (d’Orbigny). Rotalina canariensis d’Orbigny, 1839, in Barker-Webb and Berthelot, Hist. Nat. lies Canaries, II, “ Foramini- feres ”, p. 130, pi. 1, figs. 34-36. Globorotalia canariensis (d’Orbigny). Galloway and Morrey, 1929, p. 25, pi. 3, fig. 11. Frequent in sample 45. (313) Globorotalia menardii (d’Orbigny). Rotalina menardii d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 273, no. 26. Globorotalia menardii (d’Orbigny). Cushman, 1927, p. 175. Rare in sample 45 (1 specimen). Genus Cancris Montfort, 1808. (314) Cancris aurieulus (Fichtel and Moll). Nautilus auricula Fichtel and Moll, 1803, Test. Micr., var., a p. 108, pi. 20, figs. a-c. Cancris aurieulus (Fichtel and Moll). Cushman and Todd, 1942, p. 74, pi. 18, figs. 1-11, pi. 23, fig. 6 (synonymy). Not uncommon in the samples of groups 1, 2 and 3. Genus Eponides Montfort, 1808. (315) Eponides procerus (Brady). Pulvinulina procera Brady, 1879 etc., (1881), Quart. Journ. Micr. Sci. XXI, p. 66 ; 1884, p. 698, pi. cv, figs, la-c ; Heron-Alien and Earland, 1914 etc. (1915), p. 717. Rare in sample 45 (two specimens). (316) Eponides repandus (Fichtel and Moll). Nautilus repandus Fichtel and Moll, 1803, Test. Micr., p. 35, pi. 3, figs. a-d. Eponides repandus (Fichtel and Moll). Montfort, 1808, 1, p. 127, 52e genre. Rare in sample 10. FOBAJVUNTFERA — COLTJNS 409 (317) Eponides tubuliferus (Heron- Allen and Earland). Truncatulina tubulifera Heron-Alien and Earland, 1914 etc. (1915), Trans. Zool. Soc. London, XX, pt. 17, p. 710, pi. lii, figs. 37—40 ; Cushman, 1924, p. 38, pi. 11, fig. 8. This small biconvex form with its peculiar development of tubules grouped within a crater-like tubercle on each dorsal chamber wall is rare in the samples of group 3. It has a wide Indo-Pacific distribution. Genus Gyroidinoides Brotzen, 1948. (318) Gyroidinoides soldanii (d'Orbigny). Gyroidina soldani d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 278, no. 5. Four specimens from sample 45 are small in size but otherwise typical of this wide- ranging species. Subfamily Glo botruxcaninae . Genus Truncorotalia Cushman and Bermudez, 1949. (319) Truncorotalia truncatulinoides (d’Orbigny). Rotalina truncatulinoides d’Orbigny, 1839, in Barker-Webb and Berthelot, Hist. Nat. lies Canaries, II, “ Foraminiferes,” p. 132, pi. ii, figs. 25-27. Globorotalia truncatulinoides (d’Orbigny). Cushman, 1927, p. 176. G. ( Truncorotalia ) truncatulinoides (d’Orbigny). Cushman and Bermudez, 1949, p. 35. Rare in sample 45 (one specimen). Subfamily Epistomininae. Genus Hdglundina Brotzen, 1948. (320) Hdglundina elegans (d’Orbigny). Rotalia (Turbinulina) elegans d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 276, no. 54. Epistomina elegans (d’Orbigny). Cushman, 1918 etc. (1931), p. 66, pi. 13, figs. 6«-c. Hdglundina elegans (d’Orbigny). Brotzen, 1948, p. 92. This well-known and widely-distributed species was common in sample 45, but not found in the shallower-water samples. Genus Rolshausenia Bermudez, 1952. (321) Rolshausenia inflata (Millett). Rotalia schroeteriana Parker and Jones, var. inflata Millett, 1898 etc. (1904), Journ. Roy. Micr. Soc., p. 504, pi. x, figs. 5a-c ; Heron-Alien and Earland, 1914 etc. (1915), p. 719. Specimens close to Millett’s figure were not uncommon in the samples of groups 1, 2 and 3. Bermudez (1952) refers this form to his new genus Rolshausenia. Genus Torresina Parr, 1947. (322) Torresina haddoni Parr. Torresina haddoni Parr, 1947a, Journ. Roy. Micr. Soc. LXIV, p. 133, pi. 1, figs. 4, 5 a-c, 6a-c, text-fig. 3. Three specimens of this recently-described species were found in sample 45. The largest, measuring 0-83 mm. in its greater diameter, showed on its ventral face the structure 410 GREAT BARRIER REEF EXPEDITION described by Parr as a “ lunate apertural chamberlet As only one specimen showing this development was found, it was not examined by sectioning or breaking. However, examination in a liquid medium suggests that the “ chamberlets ” are in fact only curved inflated alar extensions of the ventral chamber wall with a deep sinus or re-entrant, basically similar to those seen in other discorbid genera. The two smaller specimens measured 0-41 and 0-32 mm. in diameter respectively. The smaller was incomplete, the other showed only a simple, rather inflated alar flap extending over the preceding whorl. This specimen compared closely with a topotype specimen from Paine Island in Parr’s collection, in which a similar structure was observed. The diameter of the proloculus of the two larger specimens was the same, ca. 0-08 mm. in diameter, that of the incomplete specimen being 0-03 mm., suggesting that it may be the microsplieric form. There would appear to be little doubt that the large specimen figured by Parr (figs. 4 and 5«-c) represents the full development of the species, as suggested by its author. Genus Pseudoparr ella Cushman and Ten Dam, 1948. (323) Pseudoparr ella pulchra (Cushman). Pulvinulinella pulchra Cushman, 1933a, Contrib. Cushman Lab. Foram. Research, 9, pt. 4, p. 92, pi. 9, figs. 10a-c. This minute form is fairly common in the finer fractions of the samples of groups 1, 2 and 3. Cushman states that it has a wide Indo-Pacific distribution. Genus Epistomaroides Uchio, 1952. (324) Epistomaroides polystomelloides (Parker and Jones). Discorhina polystomelloides Parker and Jones, 1865, Phil. Trans. CLY, p. 421, pi. xix, figs. 8 a-c ; Brady, 1884, p. 652, pi. xci, figs, la-c ; Heron-Alien and Earland, 1914 etc. (1915), p. 698, pi. liii, figs. 19-23. (?) Epistomaria polystomelloides (Parker and Jones). Howchin and Parr, 1938, p. 303, pi. xvii, figs. 5-7, 11, 13. Epistomaroides polystomelloides (Parker and Jones). Uchio, 1952, p. 158. This common Indo-Pacific tropical species occurs in most of the samples of groups 1, 2 and 3. Specimens are of normal size, averaging about 1 mm. in diameter, and have the early chambers obscured by exogenous shell-growth in the form of granules which spread in an irregular manner along the sutures and in later chambers merge into the clear glassy struts which bridge the deep sutural grooves characteristic of this species. The incidence of this exogenous growth is rather variable, specimens from the Australian region showing a good deal of granulation and up to 4 'or 5 sutural struts on the dorsal side, whereas East African specimens show little granulation and in most cases have the sutural struts confined to the periphery, possibly representing a geographical subspecies. Occasionally very large specimens are found, apparently referable to this species but having an extreme development of exogenous shell-growth. Howchin and Parr record such a specimen, ca. 4 mm. in diameter, from the Upper Pliocene of Adelaide. No such specimens have been found in the present material, but the author has one from Longreach Bay, Rottnest Island, Western Australia (collected by Mr. L. Glauert) which reaches 5 mm. in diameter. In this specimen most of the surface is covered by an anastomosing network of deep narrow ridges of clear shell substance, the coarsely perforate shell wall showing in FORAMINTFERA — COLLINS 411 the interstices. These ridges, despite their apparent irregularity, are related to the sutural grooves, arising in a complex of radial and circumferential bars immediately over the grooves, and extending circumferentially over the chamber walls. In the earlier chambers the ridges anastomose, leaving small perforate areas between them. In the later chambers the circumferential extensions of the ridges do not meet over the chamber walls, and a large perforate area is exposed. A specimen from Plum Beach, New Caledonia, in the author’s possession from the collection of the late W. J. Parr, shows intermediate characteristics, being larger than the normal (1-68 mm. in larger diameter) and having the earlier chambers obscured by peculiar flat-topped vermiculate ridges which do not anastomose. This form, however, may be distinct from the present species. The great disparity in size and comparative rarity of these larger specimens suggest that they may be the microspheric form of the species. Insufficient material is as yet available for sectioning to determine whether this is the case, or whether they are merely monstrous specimens showing gerontic characteristics. Genus Epistomariella Kuwano, 1950. (325) Epistomariella milletti, n. sp. (Plate V, figs. 9a, 6, and c.) Discorbina semimarginata Millett(not of d’Orbigny), 1898 etc. (1903), Journ. Roy. Micr. Soc.,p. 703, pi. vii, figs. 8a-c. Test trochoid and biconvex, dorsal side more convex than ventral, thin-walled, smooth, polished and finely perforate, periphery rounded and lobulate, sutures depressed. The ventral side shows inflated, roughly triangular chamberlets about two-thirds the radial height of the chamber proper, and tapering down to a deep umbilicus. The chamberlet walls are rather more densely perforate than the remainder of the test. At the peripheral j unction of the chamberlet and the chamber, the wall of the test turns in to form a groove in which there is a slit-like supplementary aperture with a slight lip, connecting to the chamberlet. Attached to the underside of the groove is a partition wall which runs diagon- ally across the chamber and is attached to the septum of the preceding chamber and the periphery of the earlier whorl. It is not attached to the apertural face of the chamber, but has a rolled and thickened free edge, leaving a narrow space next to the apertural face which gives communication between chamber and chamberlet (Text-fig. 3). On the dorsal side the chambers are inflated, and a supplementary slit-like aperture in the depressed suture communicates with the chamber. There is an indication of a narrow Text-fig. 3. — Epistomariella milletti n. sp. Specimen broken to show internal structures, a primary aperture, co communicating opening, c chamberlet, re rolled edge of partition, pg peripheral groove, p partition, s septum, sf septal foramen, sa supplementary apertures. 412 GREAT BAERIER REEF EXPEDITION aperture at the base of the apertural face, on the periphery of the earlier whorl, but it is indistinct and does not appear to be open. The septal foramen between chambers is semi- circular and placed above the base of the septum. Holotype from sample 23. Dimensions of holotype : Greatest diameter 0-29 mm., thickness 0-13 mm. Comparing the species with Millett’s figures of the form which he referred to Discorbina semimarginata d’Orbigny, it would appear to be the same. Millett states that his form “ bears a strong resemblance to Discorbina rimosa of Parker and Jones, but is devoid of the chinlcs between the chambers characteristic of that species ”. The present species does possess “ chinks ”, but they are not at all obvious and are easily overlooked. Hof leer (1953) has studied specimens from the Lutetian of the Paris Basin which he refers to D. semimarginata (as Epistomaria), and places D. rimosa Parker and Jones, hitherto regarded as the genotype by original designation of Galloway’s genus Epistomaria, in the synonymy of the former species. The original figures and descriptions of both these species are quite inadequate for the purpose of critical comparison, and it does not appear that the original material of either of the authors has been re-studied, leaving the question of the identity of D. rimosa with D. semimarginata somewhat doubtful. With this reservation, Epistomaria semimarginata (d’Orbigny), as diagnosed by Hofker, has very different characteristics from the present species, in particular a granular, more or less opaque shell wall and a lobulate toothplate structure on the ventral side, rather than actual chamberlets. The present species closely resembles the figures of Discorbina rimosa Parker and Jones, which is recorded by the authors from “ Australian coral reefs ” as well as from Grignon. Thefi description, however, mentions a thick shell wall and large pores, neither of which characteristics fit the present form. Considering the wide concept of species held by these authors, it is hard to say whether their figure or description referred to the Eocene or the Recent specimens studied by them, but it seems possible that their Australian specimens were actually the species now being dealt with, lumped with the Eocene form on the grounds of certain similarities such as the marginal apertures. If this were the case, it might explain the disparity between their figure with its smooth, apparently poreless surface and the description quoted above, if the figures were executed from recent spechnens. This could only be decided by reference to the original material. The generic position of this species presents some difficulties. Epistomariella Kuwano, 1 950 was erected for the reception of forms with a smooth, finely perforate shell wall and rudimentary chamberlets. However, the distinctions between the granular wall of Epistomo- maria (fide Hofker), the coarsely perforate wall of Epistomaroides and the finely perforate and smooth wall of Epistomariella appear to be more fundamental than the degree of develo- ment of chamberlets, and the present specimens are therefore placed in Epistomariella on this basis. Genus Anomalinella Cushman, 1927. (326) Anomalinella rostrata (Brady). Truncatulina rostrata Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 65 ; 1884, p. 668, pi. xciv, figs. 6 a-c. Anomalinella rostrata (Brady). Cushman, 1927a, p. 93. Rare in sample 40. This species has a generally Eastern Pacific distribution. FORAMIXLFERA — COLLIXS 413 Genus Siphonina Reuss, 1856. (327) Siphonina tubuhsa Cushman. i Siphonina tubulosa Cushman, 1924, Publ. 342, Carnegie Instit. Washington, p. 40, pi. 13, figs. 1, 2. This species has a wide Indo-Pacific distribution and occurs also in southern Australian waters. Rare in samples of groups 1, 2 and 3. Genus Siphoninoides Cushman, 1927. (328) Siphoninoides echinatus (Brady). PlanorbtUina echinata Brady, 1879, Quart. Journ. Micr. Sci. XIX, p. 283, pi. viii, fig. 31. T runcatulina echinata Brady, 1884, p. 670, pi. xcvi, figs. 9-14 ; Millett, 1898 etc. (1904), p. 490 ; Heron- Alien and Earland, 1914 etc. (1915), p. 711. Siphoninoides echinata (Brady). Cushman, 1927, p. 77, pi. 16, fig. 12. Rare in the samples of groups 1, 2 and 3. (329) Siphoninoides glabrus (Heron- Allen and Earland). T runcatulina glabra Heron-Alien and Earland, 1914 etc. (1915), Trans. Liun. Soc. London, Zool., XX, pt. 17, p. 711, pi. Hi, figs. 41-47. Siphoninoides glabra (Heron-Alien and Earland). Cushman, 19275, p. 13. This smooth form was rather more common than S. echinatus in the samples of group 1, 2 and 3. Subfamily Rotaliinae. Genus Rotalia Lamarck, 1804. (330) Rotalia erinacea Heron- Allen and Earland. Discorbina imperatoria (d’Orbigny), var. globosa Millett (not Nonionina globosa v. Hagenovv), 1898 etc. (1903), Journ. Roy. Micr. Soc., p. 701, pi. vii, figs. 6a-c. Rotalia erinacea Heron-Alien and Earland, 1914 etc. (1915), Trans. Linn. Soc. London, Zool. XX, pt. 17, p. 720, pi. liii, figs. 23-26. This minute species with its globose chambers and long slender spines was rare but well distributed in the shallow- water samples of groups 1, 2 and 3. The apertural charac- teristics are obscure, but it does not appear to possess the angled extension of the aperture into the apertural face, regarded by Finlay (1939) as diagnostic of the spined genus Calcarina. It is therefore left in Rotalia as described, with some doubt. (331) Rotalia murrayi Heron- Allen and Earland. Rotalia murrayi Heron-Alien and Earland, 1914 etc. (1915), Trans. Linn. Soc. London. Zool. XX, pt. 17, p. 721, pi. liii, figs. 27-34. This species has approximately the same distribution in these samples as R. erinacea. The aperture appears to open into the umbilical cavity, but is obscure, and the affinities of the species remain rather doubtful. It is therefore recorded as described. 414 GREAT BARRIER REEF EXPEDITION Genus Streblus Fischer, 1817. (332) Streblus papillosus (Brady). Rotalia papillosa Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 708, pi. civ, figs. 9 a, b, c. This typical Indo-Pacific species was fairly common in the samples of group 3. (333) Streblus convexus n. sp. (Plate Y, figs. 10a, b and c.) Test subcircular in outline, dorsal side convex, ventral side flattened, periphery rounded and slightly lobulate. Chambers on dorsal side are in 2-3 whorls with up to 11 chambers in the last whorl, sutures oblique, clear and slightly limbate. Chambers on ventral side triangular, wall finely punctate, suture splitting near periphery and leaving a tapering groove which deepens toward the umbilicus, chambers edged with a thin imperforate limbate margin, rounded at the umbilical angle and forming an apical knob of clear shell material in the earlier chambers only. Umbilicus partly filled with a large flat-topped plug, subpolygonal in shape and sparsely perforate. Aperture consists of a low slit at the umbilical margin of the last chamber. Holotype from sample 8. Occurs in the samples of groups 1, 2 and 3. Dimensions of holotype : Diameter 0-56 mm., thickness 0-36 mm. This species differs from S. beccarii (Linnaeus) in its smaller dimensions, greater convexity on dorsal side and less exogenous shell growth. (334) Streblus tepidus (Cushman). Rotalia beccarii (Linnaeus), var. tepida Cushman, 19266, Publ. 344 Carnegie Instit., Washington, p. 79, pi. 1. Streblus catesbyanus (d’Orbigny), var. tepida (Cushman). Bermudez, 1952, p. 74, pi. xii, fig. 3. Specimens have been compared with examples of Cushman’s variety from the Dry Tortugas, Florida, and are found to be inseparable. This small and delicate form, which does not exceed 0-35 mm. in diameter, is very different from the large, heavy and decorated S. beccarii from the Adriatic. Bermudez has placed it as a variety of S. catesbyanus d’Orbigny, presumably on account of its lack of an umbilical plug. Its distinctive characters, together with its apparently wide tropical distribution, are here considered to justify its specific separation. Family Ceratobuliminidae. Subfamily Ceratobuliminae. Genus Geminospira 'Makiyama and Nakigawa, 1941. (335) Geminospira bradyi Bermudez. Bulimina convoluta Brady (not B. pupoides var. convoluta Williamson), 1884, Rep Voy Chall., Zool. 9, p. 409, pi. cxiii, figs. 6a, b ; Millett, 1898 etc. (1900), p. 274, pi. ii, fig. 9. B seudobulimina convoluta Glaessner (not of Williamson), 1937, p. 23, pi. 2, figs. 26, 27 ; Cushman, 1947, p. 77 ; Hofker, 1950, p. 386, text-figs. 265, 267. Geminospira bradyi Bermudez, 1952, pi. xiii, figs. 7 a, b (no description or synonymy). Four specimens of this rare form were found in sample 45. Hofker (loc. cit.) has studied the internal structures in detail, and the author has nothing to add, except to note that in well-preserved specimens there is a clear area surrounding the apertural fold and extending FORAMINTFERA — COLLINS 415 for some distance dorsally, ventrally and toward the periphery, bordered by a densely perforated area which extends to the suture, as in the genus Cancris. This feature is sug- gested in Millett’s figure of the species. The nomenclature of this species has long been doubtful. The identity of Brady’s and Millett’s specimens from the tropical Indo-Pacific with Williamson’s North Atlantic species has been seriously questioned, although the author is unaware of any actual com- parisons being made between specimens representing the original material studied by these authors, or even specimens from the two localities concerned. With these reservations, the re-naming of Brady’s form by Dr. Bermudez appears to be well justified. The earlier relegation of this species to the genus Pseudobulimina Earland has little to justify it when comparison is made with the type species of that genus, P. chapmani Heron- Allen and Earland. The wall in G. bradyi is thin and transparent, and in P. chapmani thick and opaque, though it must be noted that Wood (1948) found that the shell wall in both species was radiate. The evolute development of the former contrasted with the tight helicoid spiral of the latter is also noteworthy, and in all the differences amount to quite as much as between some well-established genera of the Rotalidea. Genus Cushmanella D. Iv. Palmer and Bermudez. (336) Cushmanella primitiva Cushman and McCulloch. Cushmanella primitive Cushman and McCulloch, 1940, Allan Hancock Pacific exped. 6, pt. 3, p. 163, pi. 18, figs. 6-8, 10. A few specimens referable to this species were found in most of the deeper-water dredgings of groups 3 and 4, and occasionally in shallow-water samples. Generally they agree well with the original description, but show a peculiar patterning in the wall caused by irregular lobulate areas of clear shell substance which contrast with the finely-perforate character of the wall in general. The supplementary aperture in the apertural face is small, rounded and obscure, usually marked by a depression in the wall. It can, however, be seen as a rounded opening in the septal face of the penultimate chamber, where it is presumably enlarged by resorption. Genus Lamarckina Berthelin, 1881. (337) Lamarckina scabra (Brady). Pulvinulina oblonga Williamson, var. scabra Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 689, pi. cvi, fig. 8. Lamarckina scabra (Brady). Cushman, 1926c, p. 13. Bare, one specimen in sample 44. Subfamily Robertininae. Genus Robertina d’Orbigny, 1846. (338) Robertina tasmanica Parr. Robertina tasmanica Parr, 1950, B.A.N.Z.A.R. Exped. Rept. V, pt. 6. Foraminifera, p. 369, pi. xv, figs. 10a, b, 11a, b. F our specimens from sample 45 are referable to this species, described from off Tasmania, vi. 6. 28 416 GREAT BARRIER REEF EXPEDITION (339) Robertina australis n. sp. (Plate V, figs. 11a, b.) Test irregularly fusiform with inflated chambers and lobulate outline, variable in proportions, the length/width ratio varying from 1-65 : 1 to 2-55 : 1. Chambers roughly equal in height and width, up to 6 in the last whorl. Initial end subacute, oral end broadly rounded. Chamber wall thin, fragile and transparent. Aperture loop-shaped, inclined to the axis of the test, in a shallow depression in the face of the last chamber. No secondary aperture was observed. Holotype from sample 45, where it is not uncommon. Dimensions of holotype : Length 0-39 mm., width 0-20 mm. This species has the slender proportions of R. subteres (Brady), in some specimens at any rate, but differs in the inflation of the chambers and the lobulate outline. Genus Robertinoides Hoglund, 1947. (340) Robertinoides oceanicus (Cushman and Parker). Robertina oceanica Cushman and Parker, 1947, U.S. Geol Survey Prof. Paper 210-D, p. 75, pi. 18, fig. 18. One specimen was found in sample 45. The presence of a second aperture in the sutural junction of the last and penultimate chambers is clearly shown in the type figure and is apparent in the specimen studied. As only one specimen was available, the internal features could not be effectively studied, but the external features place the species in the genus Robertinoides. (341) Robertinoides subcylindricus (Brady). Bulimina subcylindrica Brady, 1879 etc. (1881), Quart. Journ. Micr. Sci. XXI, p. 56 ; 1884, p. 404, pi. 1, figs. 16a, b. Robertina subcylindrica (Brady) Cushman and Parker, 1947, p. 75, pi. 18, fig. 13 (synonymy). One small specimen from sample 45 has the long embracing early chambers of this species. Brady’s figures do not show the accessory aperture at the junction of the last 2 chambers. It is, however, clear in the present specimen and others in the author’s posses- sion. Cushman and Parker mention “ supplementary apertures very inconspicuous ” in their diagnosis. The presence of the accessory aperture in this well-marked species places it in the genus Robertinoides. Fmily Anomalinidae. Subfamily Anomalininae. Genus Anomalina d’Orbigny, 1826. (342) Anomalina tasmanica Parr. Anomalina tasmanica Parr, 1950, B.A.N.Z.A.R. Exped. Rept. V, pt. 6, Foraminifera, p. 363, pi. xiv, figs. 4o-c. Not uncommon in sample 45. This is another example of the general relationship of the deeper-water assemblage to that of south-east Australia. FORAMIXIFERA— COLLINS 417 (343) Anomalina colligera Chapman and Parr. Anomalina ammonoides Brady (not Rosalind ammonoides Reuss), 1884, Rep. Voy. Chall., Zool. 9, p. 672, pi. xciv, figs. 2, 3. A. colligera Chapman and Parr, 1937, Aust. Antarctic Exped. 1911-14, Sci. Repts., ser. C, 1, (2), p. 117, pi. ix, fig. 26 ; Parr, 1950, p. 362, pi. xv, figs. 3a-c. This widely-distributed form is not uncommon in sample 45. Subfamily Cibicidinae. Genus Cibicides Montfort, 1808. (344) Cibicides praecindus (Karrer). Rotalia praecincta Karrer, 1868, Sitzungsb. d. k. Ak. Wiss. Wien, lvii, p. 189, pi. v, fig. 7. Truncatulina praecincta (Karrer). Brady, 1884, p. 667, pi. xcv, figs. 1-3. Specimens agreeing very well with Brady's figures are not uncommon in the samples of group 3. The species has been placed in Eponides, but the present specimens show the aperture continuing along the dorsal inferior border of the chamber as in typical Cibicides. (345) Cibicides subhciidingeri Parr. Truncatulina haidingeri Brady (not Rotalia haidingeri d’Orbigny) 1884, Rep. Voy. Chalk, Zool. 9, p. 663, pi. xcv, fig. 7. Cibicides subhaidingeri Parr, 1950, p. 364, pi. xv, figs. la-c. Common in sample 45. (346) Cibicides lobatulus (Walker and Jacob). Nautilus lobatulus Walker and Jacob, 1798, Adams Essays (Kanmacher’s Edn.), p. 642, pi. xiv, fig. 36. Cibicides lobatulus (Walker and Jacob). Cushman, 1918 etc. (1931), p. 118, pi. xxi, fig. 3. Occasional in the samples of groups 1, 2 and 3, common in sample 26. Family Amphisteginidae. Genus Amphistegina d’Orbigny, 1926. (347) Amphistegina radiata (Fichtel and Moll). Nautilus radiatus Fichtel and Moll, 1798, Test. Micr., p. 58, pi. 8, figs. a-d. Amphistegina radiata (Fichtel and Moll). Hofker, 1927, p 76, pis. xxix, xxx, figs. 2, 3, 4, 6 and 7 (synonymy). One of the commonest larger foraminifera in the shallow- water samples of groups 1, 2 and 3. Family Calcarinidae. Genus Calcarina d’Orbigny, 1826. The interpretation of this genus is still a matter of disagreement amongst systematists, as to whether it should apply to the “ Rotalia ” calcar and R. venusta group or to the more complex rotalids of the spengleri group. The author accepts the views of Finlay (1939) 418 GREAT BARRIER REEF EXPEDITION based on Galloway’s 1933 classification, that Calcarina calcar d’Orbigny is the genotype of this genus, and that the spengleri group is correctly placed in Montfort’s genus Tinoporus with T. baculatus Montfort as its genotype. (348) Calcarina calcar d’Orbigny. Calcarina calcar d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 276, no. 1, Modeles no. 34. Rotalia calcar (d’Orbigny). Brady, 1884, p. 709, pi. cviii, figs. 3, 4. Calcarina calcar d’Orbigny. Finlay, 1939, p. 526. Common in practically all the shallow- water samples of groups 1, 2 and 3. (349) Calcarina venusta (Brady). Rotalia venusta Brady, 1884, Rep. Yoy. Chall., Zool. 9, p. 708, pi. ciii, figs. 2 a-c. Calcarina venusta (Brady). Finlay, 1939, p. 524. Frequent in the samples of group 3, but not so common or well- distributed as C. calcar. Some specimens develop short spines on the periphery, one to each chamber, con- firming further the generic relationship to C. calcar as adduced by Finlay from apertural characteristics. Genus Tinoporus Montfort, 1808. (350) Tinoporus baculatus Montfort. Tinoporus baculatus Montfort, 1808, Concbyl. Syst. 1, p. 46, 37e genre. Calcarina baculatus (Montfort). Cusbman, 1919a, pp. 363-368, pi. 1, fig. 3. Specimens closely following Cushman’s diagnosis of this species are not uncommon in the samples of groups 1, 2 and 3. Finlay (1939) has shown that Tinoporus Montfort is a valid genus, despite inadequacies and discrepancies in description and figuring, since it is based on the present species (Cushman 1919), giving it a clearly recognizable type species. The use of Tinoporus for the spengleri-baculatus group rather than the later Calcarina d’Orbigny has the practical advantage of allowing the latter to be retained for the C. calcar group of rotalids, which are distinct from both Rotalia and Tinoporus. It may be noted that Brady’s fig. 7, pi. cxiii of the “ Challenger ” Report appears to belong to this species, rather than spengleri to which Brady refers it. (351) Tinoporus hispidus (Brady). Calcarina hispida Brady, 1876, Proc. Roy. Irish Acad., ser. 2, 2, p. 590 ; 1884, p. 713, pi. cviii, figs. 8 and 9. This well-known species occurs commonly in the samples of groups 1, 2 and 3, usually together with T. baculatus. Genus Baculogypsina Sacco, 1893. (352) Baculogypsina sphaerulata (Parker and Jones). Orbitolina sphaerulata Parker and Jones, 1860, Ann. Mag. Nat. Hist., ser. 3, VI, p. 33, no. 8. Tinoporus baculatus Carpenter (not of Montfort), 1860, p. 557, pis. xviii, xxi ; Brady, 1884, p. 716, pi. ci, figs. 4-7. Baculogypsina sphaerulata (Parker and Jones). Sacco, 1893, p. 206. Tinoporus baculatus var. florescens Chapman, 1901, p. 386, pi. 36, fig. 4. This species is found in great quantities in the coarser sediments of the littoral region and is rare or absent in deeper- water samples, whether mud or coral sand. It is one of the FORAMNIFERA— COLLINS 419 few forms with tests able to withstand the pounding of surf, and is hence recognizable as a constituent of the coarser sediments from the surf zone where few other forms survive. Wherever the species is found in abundance in these samples a small proportion of the tests are found to have flattened and furcate spines, similar to topotype specimens of Chapman’s variety florescens from the collection of the late W. J. Parr. Variation ranges from specimens with 1 or 2 spines flattened at the tips to those with groups of short radiating spines which project from the surface of the test in 3 or 4 places. The occurrence of these forms in populations of the typical species suggests that they are individual abnormalities caused by factors such as the accidental coalescence of 2 or more spines, and as such not justifying nomenclatorial distinction. Family Cymbaloporidae. Genus Cymbaloporetta Cushman, 1928. (353) Cymbaloporetta bradyi (Cushman). Cymbalopora poeyi (d’Orbigny) var., Brady, 1884, p. 637, pi. cii, fig. 14. C. bradyi Cushman, 1924, Publ. 342, Carnegie Instit. Washington, p. 34, pi. 10, figs 2-4. Common in all groups. (354) Cymbaloporetta squammosa (d’Orbigny). Rotalia squammosa d’Orbigny, 1826, Ann. Sei. Nat. VII, p. 272, no. 8. Cymbaloporetta squammosa (d’Orbigny). Cushman, 1928a, p. 7. Common in all groups. Genus Cymbaloporella Cushman, 1927. (355) Cymbaloporella tabellaeformis (Brady). Cymbalopora tabellaeformis Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 727, pi. cii, figs. 15-18. Cymbaloporella tabellaeformis (Brady). Cushman, 1927a, p. 81, pi. 17, fig. 7. Occasional in all groups. Genus Tretomphalus Moebius, 1880. (356) Tretomphalus milletti (Heron-Alien and Barland). Cymbalopora bulloides Millett (not of d’Orbigny), 1898 etc. (1903), Journ. Roy. Micr. Soc., p. 697, pi. vii, fig. 4. C. milletti Heron-Alien and Earland, 1914 etc. (1915), p. 689, pi. li, figs. 32-35. Tretomphalus milletti (Heron-Alien and Earland). Cushman, 1934, p. 36, pi. 11, fig. 4. Fairly common in sample 24, occasional in other samples. (357) Tretomphalus planus Cushman. Tretomphalus bvlloides (d’Orbigny) var. plana Cushman, 1924, Publ. 342, Carnegie Instit. Washington, p. 36, pi. 10, fig. 8. T. planus Cushman, 1934a, p. 94, pi. 11, figs, lla-c, pi. 12, figs. 18-22. Occasional specimens in all groups. 420 GREAT BARRIER REEF EXPEDITION (358) Tretomphalus clams Cushman. Tretomphalus clarus Cushman, 1934a, Contrib. Cushman Lab. Foram. Research, 10, pt. 4, p. 99, pi. 11, figs. 6a-c, pi. 12, figs. 16, 17. The Discorbis- stage of this species was not uncommon. Some fine specimens with float-chambers intact were found in sample 27. Family Elphidiidae. In this report the genera Nonion and Nonionella have been treated as belonging to the family Chilostomellidae on the grounds of likeness in wall structure, following the work of Wood (1948). This requires a change in the family name Nonionidae as used in Bermudez’ classification of the Rotalidea, which has otherwise been followed. It appears preferable to base any such changed family name on the oldest-established and most typical genus, Elphidium. Genus Elphidium Montfort, 1808. (359) Elphidium advenum (Cushman). Polystomella subnodosa Brady (not of Montfort), 1884, p. 734, pi. cx, figs, la, b ; Chapman, 1901, p. 203 ; Millett, 1898 etc. (1904), p. 604 ; Heron-Alien and Earland, 1914 etc. (1915), p. 753. P. advena Cushman, 1922, Publ. 311, Carnegie Instit. Washington, p. 56, pi. 9, figs. 11, 12. Elphidium advenum (Cushman), 1918 etc. (1930), p. 25, pi. 10, figs. 1, 2. This common Indo-Pacific species was found in most of the samples of groups 1, 2 and 3. (360) Elphidium craticulatum (Fichtel and Moll). Nautilus craticidatus Fichtel and Moll, 1798, Test. Micr., p. 51, pi. 5, figs. h~k. Polystomella craticulata (Fichtel and Moll). d’Orbigny, 1826, p. 284, no. 3. Elphidium craticulatum (Fichtel and Moll). Cushman, 1932 etc. (1933), p. 48, pi. 11, figs. 5a, b. This large coral-reef species was common in the coarser fractions of the samples of groups 1, 2 and 3. (361) Elphidium crispum (Linne). Nautilus crispus Linne, 1758, Syst. Nat., ed. 10, p. 709. Polystomella crispa (Linne). Lamarck, 1816 etc. (1822), p. 625. Elphidium crispum (Linne). Cushman, 1939, p. 50, pi. 13, figs. 17-21 (synonymy). Large well-developed specimens were common in the samples of groups 1, 2 and 3. (362) Elphidium hispidulum Cushman. Elphidium hispidulum Cushman, 1936a, Contr. Cushman Lab. Foram. Research, 12, pt. 4, p. 83, pi. 14, 14, figs. 13a, b. Frequent in most of the samples of groups 1, 2 and 3. It was described from Western Australia, and appears to be confined to the warmer coastal waters of Australia. (363) Elphidium sp. aff. josephinum (d’Orbigny). (Plate V, fig. 12.) aff. Polystomella josephina d’Orbigny, 1846, For. Foss, bassin Tert. Vienne, p. 130, pi. 6, figs. 25, 26. Elphidium josephinum (d’Orbigny). Cushman, 1939, p. 43, pi. 11, fig. 15 (synonymy). FORAMINIFERA— COLLINS 421 One specimen from sample 9 is very close to Cushman’s figure of this species. Records are confined to the Miocene of central and southern Europe, but it is not unusual for species originating in this region or their close congeners to be found living in the Indo-Pacific. (364) Elphidium limbatum (Chapman). Polystomella macella (Fichtel and Moll) var. limbata Chapman, 1909, Journ. Quekett Micr. Club, ser, 2, 10, p. 142, pi. 10, figs. 9a, b. Elphidium macellum (Fichtel and Moll) var. limbata (Chapman). Cushman, 1932 etc. (1933), p. 50, pi. 11, figs. 9a, b. Specimens are closely similar to those from the type locality, Port Phillip Bay, Victoria. It occurs there together with E. 'macellum s. str., and therefore cannot be regarded as a geographical subspecies. The characters are well marked, and it has a wide Indo-Pacific range, being recorded by Cushman from Fiji. It is therefore treated as a distinct species in this report. Specimens are not uncommon in the samples of groups 1, 2 and 3. (365) Elphidium oceanicum Cushman. Elphidium oceanicum Cushman, 1932 etc. (1933), U.S. Nat. Mus. Bull. 161, pt. 2, p. 52, pi. 12, figs. 7a, b. This small species is fairly rare, 1 or 2 specimens per sample being found in many of the samples of groups 1, 2 and 3. (366) Elphidium pacificum n. sp. (Plate V, fig. 13.) Elphidium milletti Cushman (not Polystomella milletti Heron-Alien and Earland), 1932 etc. (1933), U.S. Nat. Mus. Bull. 161, pt. 2, p. 49, pi. 11, figs. 8a, b. Three specimens very close to Cushman’s figure of a form from the Tonga Islands were found in sample 45. Compared with the original figures of E. milletti from the Iverimba Archipelago and with East African specimens which are clearly referable to this species, the present form is seen to differ widely. The outline is much more lobulate and irregular than in E. milletti , and the costae, instead of being low or broken and beaded, are high and ridged, tending to obscure the sutures. The rounded periphery and slightly angled form of the costae are the only real points of similarity between the two forms, and it is considered that they are specifically distinct. Family Globigerinidae. Subfamily Globigerininae. Genus Globigerina d’Orbigny, 1826. (367) Globigerina bulloides d’Orbigny. Globigerina bulloides d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 277. no. 1. ModMes no. 17 and no. 76. Occasional specimens of this wide-ranging species were found in all groups. 422 GREAT BAERIEE REEF EXPEDITION (368) Globigerina sp. cf. dubia Egger. cf. Globigerina dubia Egger, 1857, Neues Jahrb. fur. Min. etc., p. 281, pi. ix, figs. 7.9 ; Brady, 1884, p 595, pi. lxxix, figs. 17a-c. Rare in sample 39. This form, which is low-spired, is referred doubtfully to G. dubia, rather than the Pacific species G. eggeri Rhumbler. Genus Globigerinella Cushman, 1927. (369) Globigerinella aequilateralis (Brady). Globigerina aequilateralis Brady, 1879, Quart. Journ. Micr. Sci. XIX, p. 71 ; 1884, p. 605, pi. lxxx, figs. 18-21. Globigerinella aequilateralis (Brady). Cusliman, 1927a, p. 87. Eare in sample 45 (1 specimen). Genus Globigerinoides Cushman, 1927. (370) Globigerinoides sacculiferus (Brady). Globigerina sacculifera Brady, 1877, Geol. Mag., Dec., II, IV, p. 535 ; 1884, p. 604, pi. lxxx, figs. 11-17, pi. lxxxii, fig. 4. Globigerinoides sacculifera (Brady). Cushman, 1927, p. 173. Common in sample 45. (371) Globigerinoides conglobatus (Brady). Globigerina conglobata Brady, 1879, Quart. Journ. Micr. Sci. XIX, p. 72 ; 1884, p. 603, pi. lxxx, figs. 1-5, pi. lxxxii, fig. 5. Globigerinoides conglobata (Brady). Cushman, 1927a, p. 173. Occasional specimens were found in all groups. Subfamily Pulleniatininae. Genus Pulleniatina Cushman, 1927. (372) Pulleniatina obliquiloculata (Parker and Jones). Pullenia obliquiloculata Parker and Jones, 1865, Phil. Trans. CLV, pp. 368, 421, pi. xix, fig. 4. Pulleniatina obliquiloculata (Parker and Jones). Cushman, 1927a, p. 90, pi. 19, fig. 5. Rare in sample 45 (2 specimens). Genus Sphaeroidinella Cushman, 1927. (373) Sphaeroidinella dehiscens (Parker and Jones). Sphaeroidina dehiscens Parker and Jones, 1865, Phil. Trans. CLV, p. 369, pi. xix, figs. 5a, b. Sphaeroidinella dehiscens (Parker and Jones). Cushman, 1927a, p. 90, pi. 19, fig. 2. One typical specimen was found in sample 45. FORAMEXIFERA — COLLIXS 423 Family Pegidiidae. Genus Physalidia Heron- Allen and Earland, 1928. (374) Physalidia ? reniformis (Heron- Allen and Earland). Discorbina reniformis Heron-Alien and Earland, 1914 etc. (1915), Trans. Zool. Soc. London, XX, pt. 17, p. 696, pi. lii, figs. 7-14. Two specimens were found in sample 29. The position of this rarely-recorded species is doubtful, though it definitely does not belong in Discorbis s. str., and is probably not closely related to that genus. It is here referred with some doubt to the genus Physalidia because of its ovoid shape, its thick, pitted and coarsely-perforate shell wall, and its obscure aperture. This genus was described as having opposed chambers, but the figures of some species (cf. P. earlandi Bermudez) could be interpreted as showing loose spiral growth. The present species is trochospiral, tending to open out in the last 1 or 2 chambers. Taking all the observed characters into account, its affinities are probably with the Pegidiidae. Genus Pegidia Heron- Allen and Earland, 1928. (375) Pegidia dubia (d’Orbigny). Rotalia dubia d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 274, no. 24. Pegidia dubia (d’Orbigny). Heron-Alien and Earland, 1928, Journ. Roy. Micr. Soc., p. 294. Two characteristic specimens of this rarely-recorded species were found, one each in samples 35 and 40. Family Planorbulinidae. Subfamily Planorbulininae. Genus Planmbidina d’Orbigny, 1 826. (376) Planorbulina rubra d’Orbigny. Planorbulina rubra d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 200, no. 4 ( nomen nudum). Gijpsina rubra (d’Orbigny). Heron-Alien and Earland, 1914 etc. (1915), p. 725, pi. liii, figs. 35-37. Planorbulina rubra d’Orbigny. Parr, 1932, p. 232, pi. xxii, figs. 51a-c ; 1950, p. 368. Parr has recorded this species from various localities on the Australian coast. Large incrusting specimens were found in some of the samples of groups 2 and 3, some with a faint pink tinge, but mostly white. (377) Planorbulina acervalis Brady. Planorbulina acervalis Brady, 1884, Rep. Voy. Chall., Zool. 9, p. 637, pi. xlii, fig. 4. Occasional specimens were found in most of the samples of groups 1, 2 and 3. Genus Acervulina Schultze, 1854. (378) Acervulina inhaerens Schultze. Acervulina inhaerens Schultze, 1854, Organ, des Polythal., p. 68, pi. vi, fig. 12. Gypsina inhaerens (Schultze). Brady, 1884, p. 718, pi. cii, figs. 1-6. Acervulina inhaerens Schultze. Cushman, 1918 etc. (1931), p. 134, pi. xxv, fig. 2 ; Parr, 1950, p. 368. Occasional specimens were found in most of the samples of groups 1, 2 and 3. 424 GREAT BARRIER REEF EXPEDITION Subfamily Gypsininae. Genus Gypsina Carter, 1877. (379) Gypsina vesicularis (Parker and Jones). Orbitolina vesicularis Parker and Jones, 1860, Ann. Mag. Nat. Hist., ser. 3, 6, p. 31, no. 5. Gypsina vesicularis (Parker and Jones). Carter, 1877a, p. 173. This common coral-reef species occurred in most of the samples of groups 1, 2 and 3. Particularly well-preserved specimens were found in sample 38. (380) Gypsina fimbriata (Chapman). (Plate Y, figs. 14a, b.) Planorbulina acervalis Brady var. fimbriata Chapman, 1900a, Journ. Linn. Soc., Zool. XXVIII, no. 181, p. 143, pi. 20, fig. 4. Six specimens were found in sample 38. The test is circular, slightly concave below, campanulate above, with raised rounded centre and flattened periphery. The chambers of the body of the test are subpolygonal, bordered by a clear imperforate wall with papillose prominences at the junctions of polygons, and filled in with a thin perforate diaphragm, added alternately in adjacent chambers so that the surface of the test is on two levels. These chambers are slightly smaller than those of G. vesicularis which occurs in the same sample, but are otherwise clearly similar, and bear little resemblance to the inflated, coarsely perforate chambers of Planorbulina acervalis. Fringing the periphery is an alternating series of elongate, subrectangular chambers, with imperforate wall and perforate diaphragm, differing only in size and shape from the other chambers. These fringing chambers occasionally anastomose or fork at the extremity, and are normally separated by a space which will be filled by the next-formed series. The upper surface of the test is brownish, the fringing chambers and the surface layer on the underside being glass-clear. Diameter ca. 1 mm. From Chapman’s figures and description, this is evidently the form he had from Funa- futi. At first sight the fringing series of chambers suggests relationship with Planorbulinella, but there are no special apertures in these chambers, and in all other respects the species is a typical Gypsina. Chapman’s illustrated specimen appears somewhat eroded, which may account for its reference to Planorbulina. Genus Sphaerogypsina Galloway. (381) Sphaerogypsina globulus (Reuss). Ceriopora globulus Reuss, 1847, Haidinger’s Naturw. Abhandl. II, p. 33, pi. v., fig. 7. Gypsina globulus (Reuss). Brady, 1884, p. 717, pi. cl., fig. 8. Sphaerogypsina globulus (Reuss). Galloway, 1933, p. 309, pi. 28, figs. 13, 14. Occurs with G. vesicularis in the samples of groups 1, 2 and 3, with particularly good specimens in sample 38. FORAMINIEERA— COLLINS 425 Genus Planogypsina Bermudez, 1952. (382) Planogypsitui squamiformis (Chapman). Gypsina vesicularis Parker and Jones var. squamiformis Chapman, 1901, Journ. Linn. Soc., Zool. XXVIII, p. 200, pi. 19, figs. 15a, 6. Planogypsina squamiformis (Chapman). Bermudez, 1952, p. 124, pi. xxiv, fig. 15. One specimen from sample 38 agrees with Chapman’s diagnosis. The chambers are acervuline, in one plane, with a thin wall and slightly inflated perforate diaphragm. On the underside the trochospiral early stage, coloured brown, is clearly visible. The specimen is 1-2 mm. in diameter, but the periphery is broken all round and it is evidently part of a larger specimen. Genus Carpenteria Gray, 1858. (383) Carpenteria monticularis Carter. Carpenteria monticularis Carter; 1877b, Ann. Mag. Nat. Hist. 4, p. 14, pi. xiii, figs. 9.12 ; Brady, 1884, p. 677, pi. xcix, figs. 1-5. A few specimens, adherent to calcareous algae, were found in the samples of group 3, wherever bottom conditions favoured the growth of sessile forms, particularly in sample 29. Genus Homotrema Hickson, 1911. (384) Homotrema rubrum (Lamarck). Millepora rubra Lamarck, 1816, Hist. Nat. Anim. sans Vert., p. 202. Homotrema rubrum (Lamarck). Hickson, 1911, Trans. Linn. Soc. London, Zool., ser. 2, 14, pp. 445-447, pis. 30-32 ; Emiliani, 1951, pp. 143-147, pis. 15, 16 (synonymy). Occasional fragments were found in the coarser coral-sand residues of samples of groups 1, 2 and 3. Genus Miniacina Galloway, 1933. (385) Miniacina miniacea (Pallas) subsp. alba (Carter). Polytrema miniaceum (Pallas) var. album Carter, 18776, Ann. Mag. Nat. Hist., ser. 4, XIX, p. 213, pi. xiii, fig. 14 ; Brady, 1884, p. 721, pi. cl, figs. 2, 3. Occasional specimens were found in the coarser residues, particularly in sample 29. Family Nummulitidae. Subfamily Nummulitinae. Genus Operculina d’Orbigny, 1826. (386) Operculina ammonoides (Gronovius). Nautilus ammonoides Gronovius, 1781, Zooph. Gron. exh. Anim. Quadr., Leyden, p. 282, pi. xix, figs. 5, 6. Operculina ammonoides (Gronovius). Carpenter, Parker and Jones, 1862, p. 810 ; Chapman and Parr, 1938, p. 290, pi. xvii, figs. 12-16, text-fig. 5 (synonymy). This is the commoner of the two species present in the samples, and is frequent in most of the samples of groups 1, 2 and 3. 426 GREAT BARRIER REEF EXPEDITION (387) Operculina bartschi Cushman. 0. bartschi Cushman, 1921, U.S. Nat. Mus. Bull. 100, p. 376, text-fig. 13 ; Chapman and Parr, 1938, p. 292, pi. xvii, figs. 17-20, text-fig. 6. Material dredged from Linden Bank (samples 31, 32 and 33) yielded a series of com- planate and expanded specimens which are referred to the above species. No 31 contained three very large specimens averaging about 2 cm. in diameter and no. 33 had one broken specimen which had been in the same size range ; in no. 32 the specimens did not exceed 6 mm. An example of each type was sectioned, with the following result : Diameter of proloculus . Maximum diameter of test Number of whorls Number of chambers in last whorl Total number of chambers Small form. 0 • 04 mm. 5-3 mm. 4 27 57 Large form. 0-02 mm. 19-0 mm. 61 38 140 The specimens thus fall into the normal pattern of alternation of generation, with the microspheric form larger and having more chambers than the megalospheric form. Relative abundance could not be ascertained, as the specimens came from different samples, both of which had apparently been fractioned, no. 31 containing only very large forms, and no. 32 containing nothing larger than 5-6 mm., while no. 33 contained only one broken specimen of the larger form and two of the smaller. The external characters are those of 0. bartschi Cushman. In both forms the central portion is thickened, convex and beaded, while the later whorls are thin, complanate and comparatively smooth. Beading in the large specimens is confined to the earlier whorls, the last whorl being smooth. The smaller specimens are beaded throughout, some specimens having the more prominent beading along the sutures, others in a row along the middle of the chamber between the sutures. A strong rounded keel is present, and the larger specimens are somewhat warped out of plane. Sutures are highly recurved and somewhat wavy and irregular in the later whorls. Some specimens of the smaller form show regular thickenings of the outer face of the septum, marked also by a local widening of the septal canal and suggesting incipient chamber- let growth. They do not, however, amount to partial septation as recorded by Chapman and Parr for 0. complanata Defrance. Genus Operculinella Yabe, 1918. 388. Operculinella venosa (Fichtel and Moll). Nautilus venosus Fichtel and Moll, 1798, Test. Micr., p. 59, pi. 8, figs. e-h. Operculinella venosa (Fichtel and Moll). Cushman, 1924, p. 50, pi. 17, fig. 7 : Chapman and Parr, 1938, p. 293, pi. xlvii, figs. 21, 22, text-fig. 7 (synonymy). Well-developed specimens including the evolute microspheric form were common in many of the samples of group 3. Said (1949) has identified “ Nautilus ” venosus Fichtel and Moll as a species of Amphistegina, rather than of Operculinella as considered by Cushman and later authors. Fichtel and Moll’s fig. “/”, pi. 8, which shows only one side of the test, FORAMIXIFERA— COLLINS 427 could be taken as representing either the dorsal view of Amphistegina or one side of the symmetrical Operculinella. In the present material there are specimens, considered to be the megalospheric form of Amphistegina radiata, which correspond as closely to fig. “/ ” as do other specimens to fig. “ b ”, of “ Nautilus ” radiatus on the same plate, suggesting that venosa may be a junior synonym of A. radiata. While there is apparently some doubt of the status of Nautilus ,rvenosus, it appears to the author that there is as yet insufficient evidence to justify any re-naming of the present species. Subfamily Heterostegininae. Genus Heterostegina d’Orbigny, 1826. 389. Heterostegina suborbicularis d’Orbigny. Heterostegina suborbicularis d’Orbigny, 1826, Ann. Sci. Nat. VII, p. 305 ; Hofker, 1927, p. 70, pis. xxxv, xxxvi, figs. 3, 6-12 (synonymy). This widely-distributed species is well represented in most of the samples of group 3, but does not occur in the reef collections nor in the two deeper-water samples. 390. Heterostegina operculinoides Hofker. Heterostegina operculinoides Hofker, 1927, Foram. Siboga Exped., pt. 1, p. 67, pis. xxxiv, xxxvi, figs. 1, 2, 4, 5. Six specimens from sample 46 are referable to this species, the two largest being ca. 7 mm. in diameter. The tests are thin, complanate and evolute, only the first few chambers being thickened to form a small raised and beaded boss. The sutures of the early whorls are beaded and limbate, the beading becoming less prominent in the later whorls. The radial sutures of the chamberlets are also limbate, but not so strongly as the septal sutures, so that the surfaces of the chamberlets appear as depressed rectangular areas. The form of the chamberlets is characteristic, being long and narrow, this character being assumed in the earliest whorls. Only the form regarded by Hofker as megalospheric was found. This sample apparently derives from a plankton net which bottomed at 205 metres. In this area, therefore, H. operculinoides appears to be confined to the deeper waters of the continental shelf, whi.e H. suborbicularis represents the genus in shallower water. Genus Cycloclypeus Carpenter, 1856. (391) Cycloclypeus carpenteri Brady. Cycloclypeus carpenteri Brady, 1881, Quart. Journ. Micr. Sci. XXI, p. 67 ; Chapman, 1900, p. 22, pi. 2, figs. 6, 7, pi. 3, figs. 1-5. One specimen of the megalospheric form (= C. guembelianus Brady) was found in sample 46. The records show that this species is confined to the deeper off-shore waters, and it did not occur in any other sample. PREVIOUS WORK IN THE REGION. Up to the time of the voyage of the “ Challenger ”, comparatively little intensive study was made of the Barrier Reef foraminiferal fauna. Parker and Jones had described 428 GREAT BARRIER REEF EXPEDITION species from “ Australian coral reefs ”, and the researches of Carpenter into the detailed morphology of the larger tropical foraminifera are of considerable importance. Unfortu- nately the “ Challenger ” collections were confined to the Torres Strait islands. Neverthe- less the records then made have been very relevant to the present work, particularly those from Raine Island. Since the “ Challenger ” report, only short papers giving faunal lists or dealing with particular species have been published. Jensen (1905) listed foraminifera from Palm and Lizard Islands and described new species. Jensen and Goddard (1907) listed species from Townsville. Chapman (1898) described Haddonia from Torres Straits, (1908) recorded dimorphism in Alveolinella from Cairns Reef, and (1931) reported on bore samples from Michaelmas Reef. Chapman and Parr (1938) studied the species of Operculina and Operculi- nella occurring in the region. Cushman has described new species from the area and (1942) reported on a boring at Heron Island. Cushman and McCulloch (1940, 1942 and 1950) have described new species from West Molle Island, collected by the Allan Hancock Pacific Expeditions. Parr has from time to time and in the course of other work referred to and resolved many questions of nomenclature in the case of Barrier Reef species, and in his last work (1950) recorded many species as ranging from Tasmania to North Queensland. He erected the genus Tonesina (1947) to receive a new species from Torres Straits. In general, the foraminifera of this region have been more closely studied than those of the subtropical waters of southern Queensland and New South Wales. ZOOGEOGRAPHICAL RELATIONSHIPS. The shallow-water assemblages of the Barrier Reef, as might be expected, show a close relationship with faunas recorded for other locations in the tropical Indo-Pacific, ranging from the east coast of Africa to the islands of the central Pacific and including the Malay Archipelago. There is also a relationship with the raised beaches and shallow- water deposits of the warm phases of the Quaternary in southern Australia, when Margino- pora complanata flourished as far south as Yorke Peninsula, and other tropical forms penetrated as far as the south coast of Victoria. A further relationship is to be noticed in the tendency for shallow-water tropical species to be represented by a closely-related cool- water form in south-east Australian waters, as in the case of Discorbis subvesicularis n. sp. and the well-known D. dimidiatus Parker and Jones. Deeper-water assemblages include many species recorded from Victorian and Tas- manian waters by Chapman, Parr and others, indicating the existence of an eastern Australian continental-shelf fauna which requires much more study before its composition and north-south distribution will be fully understood. This relationship is brought out to some extent by Parr (1950) in recording species from off Tasmania and from the Great Australian Bight and noting their occurrence in Queensland waters. The present collections throw some light on the northward distribution of previously- known cool-water forms, but much more work in this direction is needed, particularly in the New South Wales and South Queensland coastal waters, from which foraminiferal records are scanty. Greater knowledge of the climatic and depth distribution of recent foraminifera in these areas is needed for the fuller understanding of such factors in relation to Quaternary palaeoecological and climatic studies in southern Australia. FORAMINIFERA— COLLINS 429 DEPTH INDICATIONS IN RELATION TO REEF BORINGS. The results of this work have been considered in regard to the light they may throw on problems of reef-building, such as those raised by the Great Barrier Reef bores, 1926 and 1937 (Reports of the Great Barrier Reef Committee, vol. v, 1942). Cushman (ibid., p. 112) suggests that where mixed assemblages of worn larger shallow- water foraminifera and smaller well-preserved deeper-water species occur, greater weight is to be given to the deeper-water specimens in determining the living fauna and hence the probable depth during deposition. Richards and Hill (ibid., p. 68) draw the inference from Cushman’s determinations that, if the above conclusions are correct, alternating shallow and deep-water conditions must have occurred in the area of these bores, and state further that no evidence has been found in favour of such a history, the evidence of the mollusca being definitely at variance with it. Later, the authors state their opinion that the deeper-water types of foraminifera recog- nized by Cushman have been washed into the shallow shore area and that their smaller size has protected them from wear. The evidence of the present collections is decidedly in favour of the latter view. Wherever conditions of deposition on the reef-flat or mangrove-pool stations have allowed fine material to collect, such mixtures have been found. The smaller foraminifera include both pelagic and benthic species, the larger forms usually represented by worn specimens of Baculogypsina, Calcarina, Marginopora and Amphistegina, perhaps the most character- istic forms found in the coarse coral sands of the surf zone. Larger foraminifera derived from dredgings a few fathoms deep in the immediate off-shore zone are usually well preserved, though eroded specimens are of course found occasionally. Where bottom sediments are coarse in nature, due to current-sorting, sessile organisms are found on the eroded tests of larger foraminifera. This is rare in littoral deposits, where wave-pounding wears down the tests before such organisms can get a hold. Erosion in the deeper-water specimens appears to have been caused by biological factors such as boring algae, since the tests tend to be rotten and fragmentary rather than smoothly worn down as in the surf zone. A rough classification of depth indications connected with the state of erosion of tests is as follows : Assemblage. (1) Larger tests only, Baculogypsina and Marginopora predomi- nant, in various stages of surface wear and smoothing-off of detail (2) Mixture of larger tests as in (1) and smaller species in good condition, with some mud (3) Larger tests only, Operculina, Heterostegina and Operculinella predominant, with a proportion rotten and fragmentary and serving as a base for sessile foraminifera and other orga- nisms (4) Large assemblages of species of various sizes, including the more delicate calcareous and larger arenaceous forms, in fine-grained matrix. Indication. Surf zone. Reef pools and areas of quiet deposition near shore. Deeper waters off-shore, subject to current-scouring. Deeper off-shore waters, mud bottom. The evidence of the present collections does not go much beyond this, but is sufficient to show that the conditions observed by Cushman in the Barrier Reef bores are more likely to indicate shallow-water or littoral deposition than deep-water deposits comparable 430 GREAT BARRIER REEF EXPEDITION with the depth of boring. The variations in assemblage with depth of bore do not neces- sarily mean varying depths of deposition. In shallow-water and littoral conditions, where the foraminiferal assemblage is largely a thanatocoenosis, current- and wave-sorting result in the accumulation of coarse or fine sediments in localities perhaps separated by only a few yards. Changes in local topography can result in similar variations occurring in vertical succession, giving very different assemblages by purely mechanical and acci- dental causes having no depth significance. To interpret such changes correctly requires consideration of the condition of tests, the presence or absence of sessile organisms and the nature of the matrix. LIST OF LITERATURE. Balkwill, F. P. and Millett, F. W. 1884. The Foraminifera of Galway. Journ. Micr. 3, pp. 19-28, 78-90, pis. i-iv. Bandy, 0. L. 1952. The genotype of Siphogenerina. Contr. Cushman Found. Foram. Res. 3, pt. 1, pp. 17-18. Batsch, A. J. G. F. 1791. Sechs Kupfertafeln mit Conchylien des Seesandes. Jena. Bermudez, P. J. 1952. Bstudio sistematico de los Foraminiferes Rotaliformes. Ministerio de Minas e Hidrocarburos, Venezuela, Boletin de Geologia, 2, No. 4, pp. 1-230, pis. 1-35. Berthelin, G. 1878. Liste des Foraminiferes receuilles dans la Baie de Bourgneuf et a Pornichet. 8vo. Nantes. Blainville, H. M. D. de. 1834. Manuel d’Actinologie ou de Zoophytologie, pp. 1-644, pis. i-c (in Atlas). Paris. Levrault. Brady, H. B. 1876. On some Foraminifera from the Loo Choo Islands. Proc. Roy. Irish Acad., ser. 2, 2, II, pp. 589-590. 1877. Supplementary Note on the Foraminifera of the Chalk (?) of the New Britain Group. Geol. Mag., Dec., II, LV, pp. 534-536. 1879 etc. Notes on some of the reticularian Rhizopoda of the “ Challenger ” Expedition. Quart. Journ. Micr. Sci., 1879, XIX, pp. 20-63, pis. iii-v ; pp. 261-299, pi. viii ; 1881, ibid, XXI, pp. 31-71. 1882. Report on the foraminifera. In Tizard and Murray, Exploration of the Faroe Channel during the summer of 1880, in H.M.S. “ Knight Errant ”, with subsidiary reports. Proc. Roy. Soc.. Edin- burgh, II (1880-82), pp. 708-717, pi. vi. 1884. Report on the foraminifera dredged by H.M.S. “ Challenger ” during the years 1873-76. “ Challenger ” Exped. Repts., Zool. 9, 2 vols., text and pis. 1890. Note on a new type of foraminifera of the family Chilostomellidae. Journ. Roy. Micr. Soc., 1890, pp. 567-571, text-fig. — — Parker, W. K. and Jones, T. R. 1870. A monograph of the genus Polymorphina. Trans. Linn. Soc., London, XXVII, (2), pp. 197-253, pis. xxxix-xlii. Brotzen, F. 1948. The Swedish Paleocene and its Foraminiferal Fauna. Sver. geol. Unders., ser. C, No. 493, Arsbok 42, No. 2, pp. 1-140, pis. 1-19, text-figs. 1-41. Buchner, P. 1940. Die Lagenen des Golfes von Neapel und der marinen Ablagerungen auf Ischia. Nova Acta Leopoldina. Abh. K. Leopoldinisch-Carolinisch Deutschen Akad. der Naturf. N.F., Bd. 9, No. 62, pp. 363-560, Taf. i-xxix. > Burrows, H. W. and Holland, R. 1895. In Jones, Parker and Brady (T. R. Jones). A monograph of the Foraminifera of the Crag, Part II. Pal. Soc. Mon., London, pp. 73-210, pis. 5-7. Carpenter, W. B. 1861. Researches on the Foraminifera. Phil. Trans. 150, pp. 535-586, pis. xvii-xxii. Parker, W. K. and Jones, T. R. 1862. Introduction to the Study of the Foraminifera. Ray Soc., London. Carter, H. J. 1877. Description of a new species of Foraminifera (Rotalia spiculotesta). Ann. Mag. Nat. Hist., ser. 4, XX, pp. 470-473, pi. xvi. 1877a. On a Melobesian form of Foraminifera (Gypsina melobesoides mihi) ; and further observations on Carpenteria monticularis. Ann. Mag. Nat. Hist., ser. 4, XX, pp. 172-176. 18776. On the locality of Carpenteria balaniformis, with Description of a new Species, and other Foraminifera found in and about Tubipora musica. Ann. Mag. Nat. Hist., ser. 4, XIX, pp. 209-219, pi. xii. FORAMINIFERA — COLLINS 431 Chapman, F. 1898. On Haddonia, a new Genus of tlie Foraminifera from Torres Straits. Journ. Linn. Soc., Zool. XXVI, pp. 452-456, pi. 28. 1900. On some Xew and Interesting Foraminifera from the Funafuti Atoll, Ellice Islands. Journ. Linn. Soc., Zool., XXVIII, pp. 1-27, pis. 1-4. 1900a. Foraminifera from the Lagoon at Funafuti. Journ. Linn. Soc., Zool., XXVIII, pp. 161-210, pis. 19, 20. 1901. On the Foraminifera collected around the Funafuti Atoll from Shallow and Moderately Deep Water. Journ. Linn. Soc., Zool., XXVIII, pp. 379-417, pis. 35, 36. 1908. On Dimorphism in the Recent Foraminifer, Alveolina boscii Defr. sp. Journ. Roy. Micr. Soc. for 1908, pp. 151-153, pis. ii, iii, text-fig. 1909. Recent Foraminifera of Victoria : Some Littoral Gatherings. Journ. Quekett Micr. Club, ser. 2, 10, pp. 117-146, pis. 9, 10. 1910. On the Foraminifera and Ostracoda from Soundings (Chiefly Deep-water) collected round Funafuti by H.M.S. “ Penguin ”. Journ. Linn. Soc., Zool., XXX, pp. 387-444, pis. 54-57. — 1915. Report on the Foraminifera and Ostracoda obtained by the F.I.S. “ Endeavour ” from the east coast of Tasmania and off Cape Wiles, South Australia. Australia, Dept. Trade and Customs, Fisheries, Zool. (Biol.) Results, 3 (1), pp. 3-51, pis. i-iii. 1931. A Report on Samples obtained by Boring into Michaelmas Reef, about 22 miles N.E. of Cairns, Queensland. Great Barrier Reef Committee, Repts. Ill, pp. 32—42, pis. ix, x. 1941. Report on Foraminiferal Soundings and Dredgings of the F.I.S. “ Endeavour ” along the Continental Shelf of the S.E. Coast of Australia. Trans. Roy. Soc. S. Aust., 65 (2), pp. 145-211, pis. vii-ix. — — and Parr, W. J. 1935. Foraminifera and ostracoda from soundings made by the trawler “ Bon- thorpe ” in the Great Australian Bight. Journ. Roy. Soc. W. Aust. 21 (1934/35) pp. 1-6, pi. 1. 1937. Aust. Antarctic Exped. 1911-14, Sci. Repts., Ser. C, Zool. and Bot. 1, part ii, Foramini- fera, pp. 1-190, pis. viii-x. 1938. Australian and New Zealand Species of the Foraminiferal Genera Operculina and Operculinella. Proc. Roy. Soc. Victoria, 50, part 1, pp. 279-299, pis. xvi, xvii. Chaster, G. W. 1892. Report upon the Foraminifera of the Southport Society of Natural Science District. Southport Soc. Nat. Sci. 1, pp. 54-71, pi. 1. Collins, A. C. 1953. Pleistocene Foraminifera from Port Fairy, Western Victoria. Mem. Nat. Mus. Victoria., Melbourne, 18, pp. 93-105, pi. 1. Costa, O. G. 1856. Paleontologia del Regno di Napoli, parti 2. Atti Accad. Pont. VII, pp. 145-378, pis. ix-xxvii. Cushman, J. A. 1910 etc. A monograph of the foraminifera of the North Pacific Ocean. Smithsonian Inst., U.S. Nat. Mus. Bull. 71, pt. 1, 1910 ; pt. 2, 1911 ; pt. 3, 1913, pt. 4, 1914 ; pt. 5, 1915, pt. 6, 1917. 1910a. New Arenaceous Foraminifera from the Philippines. Proc. U.S. Nat. Mus. 38, No. 1759, pp. 437-442, text-figs. 1-19. 1913. New Textulariidae and other Arenaceous Foraminifera from the Philippine Islands and contiguous waters. Proc. U.S. Nat. Mus. 44, No. 1973, pp. 633-638, pis. 78-80. 1918 etc. The foraminifera of the Atlantic Ocean. Smithsonian Inst., U.S. Nat. Mus. Bull. 104, pt. 1, 1918 ; pt. 2, 1920 ; pt. 3, 1922 ; pt. 4, 1923, pt. 5, 1924 ; pt. 6, 1929, pt. 7, 1930, pt. 8, 1931. 1918a. Foraminifera from Murray Island, Australia. Publ. 213, Carnegie Instit. Washington, pp. 289, 290, pi. 96. 1919. Recent foraminifera from off New Zealand. Proc. U.S. Nat. Mus. 56, No. 2302, pp. 593-640, pis. lxxiv-lxxv. 1919a. The relationships of the genera Calcarina, Tinoporus and Baculogypsina as indicated by recent Philippine material. U.S. Nat. Mus. Bull. 100, 1, pt. 6, pp. 363-368, pis. 44, 45. 1921. Foraminifera of the Philippine and adjacent seas. U.S. Nat. Mus. Bull. 100, pt. 4, pp. pp. 1-608, pis. 1-xcix. 1922. Shallow-water foraminifera of the Tortugas region. Publ. 311, Carnegie Instit. Washington, XVII, pp. 1-85, pis. 1-14. 1924. Samoan Foraminifera. Publ. 342, Carnegie Instit. Washington, pp. 3-75, pis. 1-25. 1925. Foraminifera of the Tropical Central Pacific. Bernice P. Bishop Mus. Bull. 27, pp. 121-144. 1926. Foraminifera of the genera Siphogenerina and Pavonina. Proc. U.S. Nat. Mus. 67, Art. 25, pp. 1-24, pis. 1-6. 1926a. The genus Chilostomella and Related Genera. Contr. Cushman Lab. Foram. Res. 1, pt. 4, pp. 73-80, pi. 11. vi. 6. 29 432 GREAT BARRIER REEF EXPEDITION Cushman, J. A. 19266. Recent Foraminifera from Porto Rico. Publ. 344, Carnegie Instit. Washington, pp. 73-84, pi. 1. 1926c. The genus Lamarckina and, its American species. Contr. Cushman Lab. Foram. Res. 2, pt. 1, pp. 7-14, pi. 1 (in part). 1927. Recent Foraminifera from off the West Coast of America. Bull. Scripps Instit. Oceanogr., Tech. Ser. 1, No. 10, pp. 114-188, pis. 1-6. 1927a. An outline of a reclassification of the Foraminifera. Contr. Cushman Lab. Foram. Res. 3, pt. 1, pp. 1-105, pis. 1-21. 19276. Foraminifera of the genus Siphonina and related genera. Proc. U.S. Nat. Mus. 72, Art. 20, pp. 1-15, pis. 1-4. 1928. Fistulose species of Gaudryina and Heterostomella. Contr. Cushman Lab. Foram. Res. 4, pt. 4, pp. 107-112, pi. 16. 1928a. Additional genera of the foraminifera. Contr. Cushman Lab. Foram. Res. 4, pt. 1, pp. 1-8. 1929. The genus Bolivinella and its species. Contr. Cushman Lab. Foram. Res. 5, pt. 2, pp. 28-34, pi. 5. 1931. Parrina, a new generic name. Contr. Cushman Lab. Foram. Res. 7, pt. 1, pp. 17, 18, pi. 3 (in part). 1931a. Two new foraminiferal genera from the South Pacific. Contr. Cushman Lab. Foram. Res. 7, pt. 4, pp. 78-82, pi. 10 (in part). 1932 etc. The foraminifera of the tropical Pacific collections of the “ Albatross ”, 1899-1900 : U.S. Nat. Mus. Bull. 161, pt. 1, 1932 ; pt. 2, 1933, pt. 3, 1942. 1933. Some new foraminiferal genera. Contr. Cushman Lab. Foram. Res., 9, pt. 1, pp. 21, 22. 1933a. Some new Recent Foraminifera from the Tropical Pacific. Contr. Cushman Lab. Foram. Res. 9, pt. 4, pp. 77-95, pis. 8-10. 1934. A Recent Gumbelitria ? from the Pacific. Contr. Cushman Lab. Foram. Res. 10, pt. 4, p. 105, pi. 13 (in part). 1934a. Notes on the genus Tretomphalus with descriptions of some new species and a new genus Pyropilus. Contr. Cushman Lab. Foram. Res. 10, pt. 4, pp. 79-101, pis. 11, 12, 13 (in part). 1936. New genera and species of the families Verne uilinidae and Valvulinidae and of the subfamily Virgulinidae. Sp. Publ. No. 6, Cushman Lab. Foram. Res., pp. 1-71, pis. 1-8. 1936a. Some new species of Elphidium and related genera. Contr. Cushman Lab. Foram. Res. 12, pt. 4, pp. 77-89, pi. 13 (in part), 14, 15. 1937. A monograph of the foraminiferal family Verneuilinidae. Sp. Publ. No. 7, Cushman Lab. Foram. Res., pp. i-xiii, 1-157, pis. 1-20. 1937a. A monograph of the foraminiferal family Valvulinidae. Sp. Publ. No. 8, Cushman Lab. Foram. Res., pp. i-xiii, 1-210, pis. 1-24. 19376. A monograph of the subfamily Virgulininae of the foraminiferal family Buliminidae. Sp. Publ. No. 9, Cushman Lab. Foram. Res., pp. i-xv, 1-228, pis. 1-24. 1939. A monograph of the foraminiferal family Nonionidae. U.S. Geol. Survey Prof. Paper 191, pp. 1-100, pis. 1-20. 1942. A report on samples obtained by the boring on Heron Island, Great Barrier Reef, Australia. Reports of the Great Barrier Reef Committee, V, Appx. 1, pp. 112-119, pis. 11-12. 1944. The genus Articulina and its species. Sp. Publ. No. 10, Cushman Lab. Foram. Res., pp. 1-21, pis. 1-4. 1944a. Foraminifera from the shallow water of the New England coast. Sp. Publ. No. 3, Cushman Lab. Foram. Res., pp. 1-37, pis. 1-4. 1945. The species of the subfamily Rdussellininae of the foraminiferal family Buliminidae : Contr. Cushman Lab. Foram. Res. 21, pt. 2, pp. 23-54. 1946. The genus Sigmoilina and its species. Contr. Cushman Lab. Foram. Res. 22, pt. 2, pp. 29-45, pis. 5, 6. 1946a. The genus Hauerina and its species. Contr. Cushman Lab. Foram. Res. 22, pt. 1, pp. 2-15, pis. 1 (in part), 2. 19466. Polysegmentina, a new genus of the Opthalmidiidae. Contr. Cushman Lab. Foram. Res. 22, pt. 1, p. 1, pi. 1 (in part). Cushman, J. A. and Bermudez, P. J. 1949. Some Cuban species of Globorotalia. Contr. Cushman Lab. Foram. Res. 25, pt. 2, pp. 26-45, pis. 5-8. Cushman, J. A. and Gray, H. B. 1946. Some new species and varieties of Foraminifera from the Pliocene of Timms Point, California. Contr. Cushman Lab. Foram. Res. 22, pt. 2, pp. 65-69, pi. 12. FORAMINIFERA — COLLINS 433 Cushman, J. A. and McCulloch, I. 1939. A Report on some Arenaceous Foraminifera. Allan Hancock Pacific Exped. 6, No. 1, pp. 1-113, pis. 1-12. 1940. Some Nonionidae in the collections of the Allan Hancock Foundation. Allan Hancock Pacific Exped. 6, No. 3, pp. 145-178, pis. 17-20. 1942, Some Virgulininae in the collections of the Allan Hancock Foundation. Allan Hancock Pacific Exped. 6, No. 4, pp. 181-230, pis. 21-28. 1948. The Species of Bulimina and related genera in the collections of the Allan Hancock Foundation. Allan Hancock Pacific Exped. 6, No. 5, pp. 231-294, pis. 29-36. 1950. Some Lagenidae in the collections of the Allan Hancock Foundation. Allan Hancock Pacific Exped. 6, No. 6, pp. 295-364, pis. 37-48. Cushman, J. A. and Parker, F. L. 1931. Recent Foraminifera from the Atlantic Coast of South America. Proc. U.S. Nat. Mus. 80, Art. 3, pp. 1-24, pis. 1-4. 1947. Bulimina and related foraminiferal genera. U.S. Geol. Survey Prof. Paper 210-D, pp. 55-176, pis. 15-30. Cushman, J. A. and Todd, M. R. 1941. Notes on the species of Uvigerina and Angulogerina described from the Pliocene and Pleistocene. Contr. Cushman Lab. Foram. Res. 17, pt. 3, pp. 70-78, pis. 17-20. 1942. The genus Cancris and its species. Contr. Cushman Lab. Foram. Res. 18, pt. 4, pp. 72-94, pis. 17-24. 1943. The genus Pullenia and its species. Contr. Cushman Lab. Foram. Res. 19, pt. 1, pp. 1-23, pis. 1-4. 1944. The genus Spiroloculina and its species. Sp. Publ. No. 11, Cushman Lab. Foram. Res., pp. 1-82, pis. 1-9. 1944a. Species of the genera Nodopthalmidium, Nodobacularietta and Vertebralina. Contr. Cushman. Lab. Foram. Res. 20, pt. 3, pp. 64-77, pis. 11-12. Cushman, J. A. and Ozawa, Y. 1930. A Monograph of the Foraminiferal Family Polymorpliinidae, Recent and Fossil. Proc. U.S. Nat. Mus. 77, Art. 6, pp. 1-185, pis. 1-40. Douville, H. 1906. Les calcaires a Fusulines de l’lndochine. Bull. Soc. Geol. France, ser. 4, 6, pp. 576-587, pis. xvii, xviii. Earland, A. 1933. Foraminifera ; Part II, South Georgia. “ Discovery ” Repts., Cambridge, VII, pp. 27-138, pis. i-iv. 1934. Foraminifera ; Part III, The Falklands Sector of the Antarctic (excluding South Georgia). “ Discovery ” Repts., Cambridge, X, pp. 1-208, pis. i-x. Egger, J. G. 1857. Die Foraminiferen der Miocan-Schichten bei Ortenburg in Nieder-Bayern. Neues Jahrb. fur Min., Jahrg. 1857, pp. 266-311, pis. 5-15. 1893. Foraminiferen aus Meersgrundproben, gelothet von 1874 bis 1876 von S. M. Sch. “ Gazelle ”. Abhandl. kon. bay. Akad. Wiss., Miinchen, Cl. II, 18, Abt. 2, pp. 195-458, 21 pis., 1 map. Ehrenburg, C. G. 1843. Verbreitung und Einfluss des mikroscopischen Lebens in Sud- und Nord-Amerika. K. Akad. Wiss., Berlin, Physik Abh., Jarhrg., 1841, pp. 291-445, pis. i-iv. 1854. Mikrogeologie. Leipzig, 2 vols. folio, 40 pis. Eimer, G. H. T. and Fickert, C. 1899. Die Artbildung und Verwandschaft bei den Foraminiferen. Zeitschr. Wiss. Zool., Leipzig, 65, pp. 599-708, text-figs, i-xlv. Emiliani, C. 1951. On the species Homotrema rubrum (Lamarck). Contr. Cushman Found. Foram. Res. II, pt. 4, pp. 143-147, pis. 15-16, fig. 1. Fichtel, L. v. and Moll, J. P. C. v. 1798. Testacea microscopica aliaque minuta ex generibus Argonauta et Nautilus ad naturam delineata et descripta. Vienna. Finlay, H. J. 1939 etc. New Zealand Foraminifera : Key Species in Stratigraphy. No. 1, Trans. Roy. Soc. New Zealand, 68, (1939), pp. 504-533, pis. 68, 69 ; No. 2, 1. c., 69, (1939), pp. 89-128, pis. 11-14 ; No. 3, 1. c. , pp. 309-329, pis. 24-29 ; No. 4, 1. c., 69, (1940), pp. 448-472, pis. 62-67. Flint, J. M. 1899. Recent Foraminifera. A descriptive catalogue of specimens dredged by U.S. Fish Commission Steamer “ Albatross ”. U.S. Nat. Mus., Ann. Rept., 1897, pt. 1, pp. 249-349, pis. i-lxxx. Fornasini, C. 1900. Le Polymorphine e le Uvigerine fossili d’ltalia. Boll. Soc. Geol. Ital. 19, pp. 132-172, 7 fiSs- 1902. Intorno a la nomenclatura di alcuni nodosaridi neogenici italiani. Mem. Accad. Sci. Istit. Bologna, 9 (1901), pp. 45-76, 27 text-figs. Forskal, P. 1775. Descriptiones Animalium, etc. 4to, Copenhagen. Frizzell, D. L. and Schwartz, E. 1950. A new Lituolid Foraminiferal Genus from the Cretaceous, with an emendation of Cribrostomella Cushman. Univ. Missouri, School of Mines and Metallurgy, Tech. Ser., Bull. 76, pp. 1-12, pi. 1, text-fig. 1. 434 GREAT BARRIER REEF EXPEDITION Galloway, J. J. 1933. A Manual of Foraminifera. 4to, Bloomington, Indiana ; pp. i-xii, 1-483, 42 pis. and Morrey, M. 1929. A Lower Tertiary Foraminiferal Fauna from Manti, Ecuador. Bull. Amer. Pal. XV, No 55, pp. 7-56, pis. i-vi. Glaessner, M. 1937. On a New Family of Foraminifera. Studies in Micropaleontology, 1, fasc. 3, pp. 19-29, pis. i, ii. 1947. Principles of Micropalaeontology. 4to, Melbourne, pp. i-xvi, 1-297, 14 pis., 7 tables, 64 text-figs. Goddard, E. J. and Jensen, H. I. 1907. Contributions to a knowledge of Australian foraminifera, Part II. Proc. Linn. Soc. New South Wales, 32, pp. 281-318, pi. vi. Gronovius, L. T. 1781. Zoophylacium Gronovianum etc. Leyden. Hada, Y. 1931. Report of the Biological Survey of Mutsu Bay. 19. Notes on the Recent Foraminifera from Mutsu Bay. Sci. Repts. Tohoku Imp. Univ., 4th Ser., Biol. YI, No. 1, pp. 45-148, text-figs. 1-95. Heron-Allen, E. and Earland, A. 1913. On some Foraminifera from the North Sea, dredged by the Fisheries Cruiser “ Goldseeker No. 3. Journ. Roy. Micr. Soc. for 1913, pp. 272-276, pi. xi. — — 1913a. Clare Island Survey, pt. 64, Foraminifera. Proc. Roy. Irish Acad. XXXI, sect. 3 (1911-15), pp. 1-188, pis. i-xiii. 1914 etc. The Foraminifera of the Kerimba Archipelago (Portuguese East Africa). Trans. Zool. Soc. London, 20, pt. 1, 1914, pp. 363-390, pis. xxxv-xxxvii ; pt. 2, 1915, pp. 543-794, pis. xl-liii. 1922. Protozoa, Part II, Foraminifera. Brit. Antarctic (Terra Nova) Exped., 1910, Nat. Hist. Repts., Zool. 6 (2), pp. 25-268, pis. i-viii. 1924. The Foraminifera of Lord Howe Island, South Pacific. Journ. Linn. Soc., Zool. 35, pp. 599-647, pis. 35-37. 1924a. The Miocene Foraminifera of the “ Filter Quarry ”, Moorabool River, Victoria, Australia. Journ. Roy. Micr. Soc., 1924, pp. 121-186, pis. 7-14. 1928. On the Pegididae, a New Family of Foraminifera. Journ. Roy. Micr. Soc. 48, pp. 283-299, 3 pis. 1932. Foraminifera. Part 1. The ice-free area of the Falkland Islands and adjacent seas. “ Discovery ” Repts., Cambridge, 4, pp. 291-460, pis. vi-xvii. Hickson, S J. 1911. On Polytrema and some allied genera. Trans. Linn. Soc. London., Zool., ser. 2, 14, 442-62, pis. 30-32. Hofker, J., 1927. The Foraminifera of the Siboga Expedition. Part I. Monograph IV, Siboga Exped., Pt. I, pp. 1-78, pis. i-xxxviii, text-figs. 1-11. 1951. The Foraminifera of the Siboga Expedition, Part III. Monograph IVa, Siboga Exped., pp. 1-508, text-figs. 1-348. 1953. The genus Epistomaria Galloway 1933, and the genus Epistomaroides Uchio 1952. Palaont. Zeitschr., 27, Nos, 3/4, pp. 129-142, text-figs. 1-14. Hoglund, H. 1947. Foraminifera in the Gullmar Fjord and the Skagerrak. Zool. Bidr. fran Uppsala, Bd. 26, pp. 3-328, pis. 1-32, text-figs. 1-312. Howchin, W. and Parr, W. J. 1938. Notes on the Geological Features and Foraminiferal Fauna of the Metropolitan Abattoirs Bore, Adelaide. Trans. Roy. Soc. South Australia, 62 (2), pp. 287-317, pis. xv-xix. Jensen, H. I. 1905. Contributions to a knowledge of Australian foraminifera ; Part I. Proc. Linn. Soc. New South Wales, 29 (1904), pt. 4, pp. 810-832, pi. xxiii. Jones, F. W. O. Rymer. 1872. On some Recent forms of Lagenae from Deep-sea Soundings in the Java Sea. Trans. Linn. Soc. London, XXX, p. 45, pi. xix. Jones, R. T. and Parker, W. K. 1860. On the Rhizopodal fauna of the Mediterranean, compared with that of the Italian and some other Tertiary deposits. Quart. Journ. Geol. Soc., XVI, pp. 292-307. and Brady, H. B. 1866 etc. A Monograph of the foraminifera of the Crag. Palaeontogr. Soc., London, pt. 1, 1866, pp. 1-72 ; pt. 2, 1895, pp. 73-210, pt. 3, 1896, pp. 211-314 ; pt. 4, 1897, pp. 315-402. Karrer, F. 1868. Die Miocene Foraminiferen-Fauna von Kostej im Banat. Sitz. Akad. Wiss. Wien, VIII, pp. 111-193, pis. i-v. Lamarck, J. P. B. A. de. 1804 etc. Explication a des planches relatives aux coquilles fossiles des environs de Paris. Ann. du Mus. (Paris), 5, 1804, pp. 179-180, 237-245, 349-357. Continued in vol. 8, 1806, pp. 383-387, pi. lxii, and vol. 9, pp. 236-240, pi. xvi. 1816 etc. Histoire naturelle des Animaux sans Vertebres. Paris, 2, 1816, pp. 193-197 ; 7, 1822, pp. 580-632. FORAMINTFERA — COLLINS 435 Lacroix, E. 1938. Revision des genre Massilim. Bull, l’lnst. Oceanogr. (Monaco), No. 754, pp. 1-11, text-figs. 1-9. Lalickeb, C. 6. and McCulloch, I. 1940. Some Textulariidae of tlie Pacific Ocean. Allan Hancock Pacific Exped. 6, No. 2, pp. 115-143, pis. 13-16. Lixne, C. A. 1758. Systema Naturae, ed. 10. 8vo, Stockholm. Loeblich, A. R. jun. and Tappan, H. 1953. Studies of Arctic Foraminifera. Smithsonian Misc. Coll. 121, No. 7, pp. 1-143, pis. 1-24, 1 text-fig. - — — 1954. Emendation of the foraminiferal genera Ammodiscus Reuss, 862, and Involutina Terquem, 1862. Journ. Washington Acad. Sci. 44, No. 10, pp. 306-310, text-figs. 1, 2. Millett, F. W. 1898 etc. Report on the recent foraminifera of the Malay Archipelago contained in anchor- mud, collected by Mr. A. Durrand, F.R.M.S. Journ. Roy. Micr. Soc., 1898, pp. 258-269, 499-513, 607-614; 1899, 'pp- 249-255, 357-365, 557-564; 1900“ pp. 6-13, 273-281, 539-549; 1901, pp. 1-11, 485-497, 619-628 ; 1902, pp. 509-528 ; 1903, pp. 253-275, 685-704 ; 1904, pp. 489-506, 597-609. 19 pis. Montague, G. 1803-8. Testacea Britannica, or natural history of British shells. 3 vols., London, 1803. Supplement (pis.) 1808. Montfort, D. de. 1808. Conchvliologie Systematicpie et Classification Methodique des Coquilles, etc. 2 vols. Paris. 1808-10. Myers, E. H. 1935. The Life History of Patellina cornigata Williamson, a foraminifer. Bull. Scripps Instit. Oceanogr., Tech. Ser. 3, pp. 355-392, pis. 10-16, 1 text-fig. Norman, A. M. 1876. In Jeffries, Preliminary Report of the Biological Results of a Cruise in LI.M.S. “ Valorous ” to Davis Strait in 1875. Proc. Roy. Soc. XXV, p. 202. 1878. On the Genus Haliphysema ,with a description of several forms apparently allied to it. Ann. Mag. Nat. Hist., ser. 5, 1, p. 265, pi. xvi. d’Orbigny, A. D. 1826. Tableau methodique de la classe des Cephalopodes. Ann. Sci. Nat., Paris., ser. 1, 7, pp. 96-314, pis. x-xvii (in Atlas). 1839. Voyage dans l’Amerique Meridional : Foraminiferes. Strasbourg, P. Bertrand, 5, pt. 5, pp. 1-86, pis. i-ix. 1839a. Foraminiferes. In Sagra, R. de la, Histoire physique, politique et naturelle de File de Cuba, pp. 1-224 (plates published separately), Paris, A. Bertrand. 18396. Foraminiferes des lies Canaries. In Barker-Webb and Berthelot, Histoire Naturelle des lies Canaries, Paris, Bethune, tome 2, pt. 2, Zool., pp. 119-146, pis. i-iii. — — — 1846. Foraminiferes fossiles du bassin tertiare de Vienne (Autriche). Paris, Gide et Cie, pp. 1-303, pis. i-xxi. Ovey, C. D. 1947. Note on the taxonomy of Svirillina limbata var. denticulata Brady. Journ. Micr. Soc. 67, pp. 17, 18. Parker, W. K. and Jones, T. R. 1860. On the nomenclature of the Foraminifera, IV, The Species enumerated by Lamarck. Ann. Mag. Nat. Hist. 6, pp. 29-40. 1865. On some Foraminifera from the North Atlantic and Arctic Oceans, including Davis Straits and Baffin’s Bay. Phil. Trans. Roy. Soc. 155, pp. 325-441, pis. xiii-xix. and Brady, H. B. 1865. On the nomenclature of the Foraminifera, XII. The Species enumerated by d’Orbigny in the Annales des Sciences Naturelles, vii, 1826, (3) The Species illustrated by Modeles. Ann. Mag. Nat. Hist. 16, pp. 15-41, pis. 1-3. Parr, W. J. 1932. Victorian and South Australian Shallow-water Foraminifera ; Part I. Proc. Roy. Soc. Victoria, XLIV (n.s.), pt. 1, pp. 1-14, pi. i ; Part II, ibid., XLIV (n.s.), pt. 2, pp. 218-234, pis. xxi, xxii. 1939. Foraminifera of the Pliocene of South-Eastern Australia. Min. Geol. Journ., Victoria, 1, No. 4, pp. 65-71, 1 pi. 1941. A New Genus, Planulinoides, and some Species of Foraminifera from Southern Australia. Min. Geol. Journ., Victoria, 2, No. 5, p. 315, text-fig. 1943. In Cotton and Godfrey, 1943, A Systematic List of the Echinodermata, Foraminifera, Hydroida and Brachiopoda of Southern Australia. Publ. No. 3, Malac. Soc. South Australia, Adelaide. 1945. Recent Foraminifera from Barwon Heads, Victoria. Proc. Roy. Soc. Victoria, 56 (n.s.), pt. 2, pp. 189-218, pis. viii-xii. 1947. The Lagenid Foraminifera and their Relationships. Proc. Roy. Soc. Victoria, 58 (n.s.), pp. 116-130, pis. vi, vii. 1947a. On Torresina, a new genus of the Foraminifera from Eastern Australia. Journ. Roy. Micr. Soc. LXIV, pp. 129-135, pi. i, text-figs. 1-3. 436 GREAT BARRIER REEF EXPEDITION Parr, W. J. 1950. Foraminifera. B.A.N.Z. Antarctic Research Exped. 1929-31, Repts., Ser. B. (Zool. and Bot.) Y, pt. 6, pp. 223-392, pis. iii-xv, text-figs. 1-8. and Collins, A. C. 1930. Notes on Australian and New Zealand Foraminifera ; No. 1, The species of Patellina and Patellinella, with a description of a new genus Annulopatellina. Proc. Roy. Soc. Victoria, 43 (n.s.), pp. 89-95, pi. iv. Reuss, A. E. 1847. In Haidinger’s Naturw. Abhandl. 2. 1850. Neue Foraminiferen aus dem Schichten des osterreichischen Tertiarsbeckens. Denkschr. Akad. Wiss. Wien, 1, pp. 365-390, pis. 46-51. 1851. Ueber die fossilen Foraminiferen und Entomostraceen der Septarienthon der Umgegend von Berlin. Zeitschr. deutsch. geol. Gesel. 3, pp. 49-92, pis. iii-vii. 1863. Beitrage zur tertiaren Foraminiferen-Fauna. Sitzungsb. d. k. Akad. Wiss. Wien, 48, pp. 36-69, pis. 1-8. 1865. Die Foraminiferen, Anthozoen and Bryozoen des Deutsches Septarienthones. Denkschr. d. k. Akad. Wiss. Wien, XXV, pp. 117-214, pis. i-xi. Rhumbler, L. 1903. Systematische Zusammenstellung der recenten Reticulosa. Arch. Protist., Jena, III, pp. 181-294. 1906. Foraminiferen von Laysan und den Chatham Inseln. Zool. Jahrb., Abt. Syst., Jena, Bd. 24, pp. 21-80, taf. ii-v. Sacco, F. 1893. Sur quelques Tinoporinae du Miocene de Turin. Soc. Beige. Geol., Pal. Hydr. Bull., Bruxelles, 7, fasc. 3. Said, R. 1949. Foraminifera of the Northern Red Sea. Sp. Publ. 26, Cushman Lab. Foram. Res. Schubert, R. 1907. Beitrage zu einer naturlichen Systematik der Foraminiferen. Neues Jahrb. fur Min., 25, pp. 233-260, text-fig. Schultze, M. S. 1854. Ueber den Organismus der Polythalamien (Foraminiferen), nebst Bermerkungen uber die Rhizopoden im allgemein. Fol., 7 pis., Leipzig. Schwager, C. 1866. Fossile Foraminiferen von Kar-Nicobar. Novara Exped. 1857-59. Wien, Bd. 2, Geol. Theil, pp. 187-268, taf. iv-vii. 1878. Nota su alcuni foraminiferi nnovi del tufo di Stretto presso Girgenti. Italy, Uff. Geol. (R. Comm. Geol. Ital.), Boll., col. 9, pp. 511-514, 519-529. Seguenza, G. 1862. Dei terreni terziari del distretto di Messina ; Parti II — Descrizione dei foraminiferi monotalamica delle marne mioceniche del distretto di Messina. Messina, T. Capra, pp. 1-84, tav. i, ii. 1880. Le Formazione Terziare nella Provincia di Reggio (Calabria). Atti Accad. Lincei, ser. 3, 6, pp. 1-446, pis. 1-17. Sidebottom, H. 1904, etc. Report on the Recent Foraminifera from the coast of the Island of Delos (Grecian Archipelago). Mem. Proc. Manchester Lit. Phil. Soc., Part I (1904), xlviii, pt. II, no. 5 ; Part II (1905), xlix, no. 5 ; Part III (1906), 1, no. 5 ; Part IV (1907) li, no. 9 ; Part V (1908), lii, no. 13 ; Part VI (1909), liii, no. 21. 1912 etc.,Lagenae of the South-west Pacific Ocean, from soundings taken by H.M.S. “ Waterwitch”, 1895. Journ. Quekett Micr. Club, ser. 2, 11, 1912, No. 70. pp. 375-434, pis. xiv-xxi. Supplementary paper, 12, 1912, No. 73, pp. 161-210, pis. xv-xviii. 1918. Report on the Recent Foraminifera dredged off the East Cost of Australia. Journ. Roy. Micr. Soc., 1918, pp. 1-25, 121-153, 249-264, pis. 1-6. Silvestri, A. 1904. Ricerche strutturali su alcune forme dei trubi di Bonfornello (Palermo). Mem. Pont. Accad. Nuovi Lincei, 22, pp. 235-276, 14 text-figs. Silvestri, O. 1872. Le Nodosarie fossile nel Terreno subappenino Italiani e viventi nei Mari d’ltalia. 11 pis., Catania. , Terquem, O. 1875 etc. Essai sur le Classement des Animaux qui vivent sur la Plage et dans les Environs de Dunkerque. 8vo, Paris, fasc. 1, 1875, pp. 1-54, pis. i-vi ; fasc. 2, 1876, pp. 55-100, pis. vii-xii ; fasc. 3, 1881, pp. 101-152, pis. xiii-xvii. 1878. Les Foraminiferes et les Entomostraces-Ostracodes du Pliocene superieur de File de Rhodes. Mem. Soc. Geol. France, ser. 3, 1, pp. 1-133, pis. 1-14. Thalmann, H. E. 1950. New names and homonyms in Foraminifera. Contr. Cushman Found. Foram. Res. 1, pts. 3 and 4, pp. 41-45. Uchio, T. 1950. Foraminiferal study of tertiary Formation near Otaki Gas Field, Chiba Prefecture (ser. 1), In Natural Gas in the Vicinity of Otaki, Chiba kin, by K. Kawai and others. Journ. Assoc. Petr. Techn. 15, No. 4, pp. 151-219, text-figs. 1-25. 1952. Foraminiferal assemblage from Hachigo Island, Tokyo Prefecture, with descriptions of some new genera and species. Japanese Journ. Geol. Geog. 22, pp. 145-159, pis. 6, 7, text-figs, a-c, table 1. FORAMINIFERA — COLLINS 437 Walker, 6. and Jacob, E. 1798. In Kanmacher, F., Adams’ Essays on the Microscope, 2nd. Edn. 4to, London. Wiesner, H. 1912. Zur Systematik adriatischer Nubecularien, Spiroloculinen, Miliolinen und Biloculinen. Arch. Protist. 25, pp. 201-239. text-figs. 1931, Die Foraminiferen. In Drygalski, E. von, Deutsche Sud-polar Exped. 1901-3. Bd. 20 (Zool. Bd. 12), pp. 53-165, Taf. i-xxiv. Williamson, W. C. 1848. On the Recent British species of the genus Laqena. Ann. Mag. Nat. Hist., ser. 2, 1, pp. 1-20, pis. i, ii. 1858. On the Recent Foraminifera of Great Britain. Ray Soc., London. Wright, J. 1886. Report on the Marine Fauna of the South-west of Ireland — Foraminifera. Proc. Roy. Irish. Acad., ser. 2, iv (1887) (Science), pp. 607-614. 1891. Report on the foraminifera obtained off the south-west coast of Ireland during the cruise of the “ Flying Falcon Proc. Roy. Irish. Acad., ser. 3, 1 (1889-91), No. 4, pp. 460-502, pi. xx. 1911. In Heron-Alien and Earland, 1908-11, On the recent and fossil foraminifera of the shore sands of Selsey Bill, Sussex. Journ. Roy. Micr. Soc., 1911, pp. 298-343. Wood, A. 1948. The structure of the wall of the test in the Foraminifera ; its value in classification. Quart. Journ. Geol. Soc. CIV, pt. 2, pp. 9-252, pis. xiii-xv. PLATE I. Fig. 1. — Discobotellina biperforata n. sp. Holotype (Form 2). “ a ” side view, “ b ” edge view. X 2. Fig. 2. — D. biperforata n. sp. Paratype (Form 1). “ a ” side view, “ b ” edge view. X 3. Fig. 3. — D. biperforata n. sp., specimen broken in half and showing internal structure. X 4. Fig. 4. — Saccammina consociata (Flint). X 28. Fig. 5. — Involutina sp. X 45. Fig. 6. — Glomospira elongata n. sp. Holotype. “ a ” side view, “ b ” end view. X 72. Fig. 7. — G. elongata n. sp., immature test showing coiling of early whorls. “ a ” side view, “ b ” end view. X 72. Fig. 8. — Lituotuba minuta n. sp. Holotype. X 72. Fig. 9. — Ammolagena clavata (Parker and Jones), two-chambered specimen. X 45. Fig. 10. — Ammomarginulina australiensis n. sp. Holotype. “ a ” side view, “ b ” apertural view. X 64. Fig. 11. — Nouria textulariformis Hada, armata n. subsp. Holotype. “ a ” side view, “ b ” apertural view. X 45. Fig. 12. — Trochammina chitinosa n. sp. Holotype. “ a ” dorsal view, “ b ” ventral view, “ c ” edge view. X 120. Fig. 13. — Gaudryina (Siphogaudryina) wrightiana Millett. “ a ” and “ b ” side views, “ c ” basal view. X 120. Fig. 14. — G. (Pseudogaudryina) concava n. sp. Holotype. “ a ” side view, “ b ” apertural view, “ c ” basal view. X 45. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. (Xat. Hist.). Reports, Yol. VI, No. 6. PLATE I. [Adlard <£• Son, Lid., Impr. PLATE II. Fig. 1. — Eggerella australis n. sp. Holotype. “ a ” side view, £; b ” apertural view. X 120. Fig. 2. — E. polita n. sp. Holotype. a ” side view, “ b ” apertural view. X 120. Fig. 3. — Dorothia inepta n. sp. Holotype. Side view. X 30. Fig. 4. — D. inepta, early chambers of another specimen exposed by grinding. X 50. Fig. 5. — Karreriella (Karrerulina) attenuata n. sp. Holotype. Side view. X 120. Fig. 6. — Quinqueloculina crassicarinata n. sp. Holotype. “ a ” and “ b ” side views, “ c ” apertural view. X 45. Fig. 7. — Q. cuvieriana d’Orbigny, queenslandica n. subsp. Holotype. “ a ” and “ b ” side views, “ c ” apertural view. X 66. Fig. 8. — Q. quinquecarinata n. sp. Holotype. “ a ” and “ b ” side views, “ c ” apertural view. X no. Fig. 9. — Massilina subrugosa n. sp. Side view. X 22. Fig. 10. — M. secans d’Orbigny, tropicalis n. subsp. Holotype. “ a ” and “ b ” side views, “ c ” apertural view. X 36. Fig. 11. — M. corrugata n. sp. Holotype. “ a ” and “ b ” side views, “ c ” apertural view. X 36. Fig. 12. — M. corrugata n. sp. Paratype. “ a ” and “ b ”, side views, “ c ” apertural view. X 36. PLATE IT. GREAT BARRIER REEF EXPEDITION 1928-29. Brit. Mus. {Nat. Hist.). Reports, Yol. VI, No. 6. [Adlard Son , Ltd.. Impr. PLATE III. Fig. 1. — Massilina minuta n. sp. Holotype. “ a ” and “ b ” side view. Fig. 2. — M. minuta, section through another specimen. X 120. Fig. 3. — S pirosigmoilina parri n. sp. Holotype. “ a ” and “ b ” side X 105. Fig. 4. — S. parri, section through another specimen. X 210. Fig. 5. — Articulina tricarinata n. sp. Holotype. “ a ” side view, “ b ” Fig. 6. — A. tricarinata, side view, x 210. Fig. 7. — A. sagra d’Orbigny. Side view. X 72. Fig. 8. — A. queenslandica n. sp. Holotype. Side view. X 120. Fig. 9. — A. queenslandica. Side view. X 120. Fig. 10. — A. queenslandica. Side view. X 120. Fig. 11. — Hauerina pacifica Cushman, rugosa n. subsp. Holotype. “ a apertural view. X 55. Fig. 12. — Triloculina littoralis n. sp. Holotype. “ a ” and “ b ” side X 85. Fig. 13. — T. quadrata n. sp. Holotype. “ a ” and “ b ” side views, “ Fig. 14. — T. sublineata (Brady) ? “a” side view, “ b ” apertural view. Fig. 15. — Nodophthalmidium gracilis n. sp. Holotype. Side view. X Fig. 16. — Cornuspirella diffusa (Heron-Alien and Earland). X 48. X 120. views, “ c ” apertural view, apertural view. X 210. ” and “ b ” side views, “ c ” views, “ c ” apertural view. c ” apertural view. X 48. X 48. 115. Brit. Mils. (X at. Hist.) GREAT BARRIER REEF EXPEDITION 1928-29. Reports, Yol. VI, No. 6. PLATE III. [Adlard <€• Son , Ltd., Imp PLATE IV. Fig. 1. — Ophthalmidium circularis (Chapman) tropicalis n. subsp. Holotype. £i a ” side view, “ b ” apertural view. X 120. Fig. 2. — Planispirinella involuta n. sp. Holotype. “ a ” side view. “ b ” apertural view. X 90. Fig. 3. — Nubeculopsis queenslandica n. gen., n. sp. Holotype. “ a ” ventral view, “ b ” dorsal view, “ c " apertural view. X 70. Fig. 4. — Fischerina pellucida Millett. X 115. Fig. 5. — Globulotuba entosoleniformis n. gen., n. sp. Holotype. “ a ” side view, “ b ” aboral view. X 115. Fig. 6. — Glandulina semistriata n. sp. Holotype. “ a ” side view, “ b ” aboral view. X 115. Fig. 7. — Buliminella latissima n. sp. Holotype. “ a ” and “ b ” side views, “ c ” apertural view. X 160. Fig. 8. — Ungulatella pacifica Cushman. X 160. Fig. 9. — Bulimina oblong a n. sp. Holotype. “ a ” and “ b ” side views, “ c ” apertural view. X 115. Fig. 10. — Globobulimina australiensis n. sp. Holotype. “ a ” side view, “ b ” apertural view, “ c ” aboral view. X 55. Fig. 11. — Mimosina rimosa Heron-Alien and Earland. X 140. Fig. 12. — Trimosina milletti Cushman, multispinata n. subsp. Holotype. Side view. X 115. Fig. 13. — Siphogenerina virgula (Brady). 11 a ” megalospheric specimen, “ b ” and “ c ” microspheric specimens. X 90. Fig. 14. — Gumbelitria vivans Cushman. X 215. Brit. Mus. (Xat. Hist.) GREAT BARRIER REEF EXPEDITION 1928-29 Reports, Yol. VI, Xo. 6. PLATE IV [At/lard <0 Son, Ltd., Itnpr. PLATE V. Fig. 1. — Bolivina alata Seguenza, jimbriata, n. subsp. Holotype. “ a ” side view, “ b ” apertural view. X 115. Fig. 2. — Loxostomurn convallarium (Millett). X 115. Fig. 3. — Bifarina elongata Millett. X 115. Fig. i.—B. queenslandica n. sp. Holotype. “ a ” side view, “ b ” apertural view. X 115. - Fig. 5. — Discorbis subvesicularis n. sp. Holotype. “ a ” dorsal view, “ b ” ventral view, “ c ” apertural view. X 75. Fig. 6. — Conorbella earlandi n. sp. Holotype. “ a ” dorsal view, “ b ” ventral view, “ c ” apertural view. X 160. Fig. 7. — Discorinopsis tropica n. sp. Holotype. “ a ” dorsal view, “ b ” ventral view, “ c ’’ apertural view. X 70. Fig. 8. — Patellinella carinata n. sp. Holotype. “ a ” side view, “ b ” dorsabview, “ c ” ventral view. X 115. Fig. 9. — Epistomariella milletti n. sp. Holotype. “ a ” dorsal view, “ b ” ventral view, “ c ” apertural view. X 115. Fig. 10. — Streblus convexus n. sp. Holotype. “ a ” dorsal view, “ b ” ventral view, “ c ” apertural view. X 55. Fig. 11. — Robertina australis n. sp. Holotype. “ a ” side view, “ b ” apertural view. X 115. Fig. 12. — Elphidium sp. aff. josephinum (d’Orbigny). X 100. Fig. 13. — E. pacificum n. sp. Side view, x 100. Fig. 14. — Gypsina Jimbriata (Chapman). “ a ” view from above, “ 6 ” edge view. X 22. Brit. Mus. (Nat. Hist.) GREAT BARRIER REEF EXPEDITION 1928-29. Reports, A~ol. AX, No. 6. PLATE A7. [Adlard <£- Son Ltd,. Inipr.